Sie sind auf Seite 1von 7

Construction and Building Materials 53 (2014) 561567

Contents lists available at ScienceDirect

Construction and Building Materials


journal homepage: www.elsevier.com/locate/conbuildmat

Evaluation of permanent deformation of asphalt mixtures using different laboratory performance tests
Tao Xu a,, Hao Wang b, Zhidong Li c, Yongli Zhao c
a

School of Civil Engineering, Nanjing Forestry University, Nanjing 210037, China Department of Civil and Environmental Engineering, Rutgers, The State University of New Jersey, NJ 08854, USA c School of Transportation, Southeast University, Nanjing 210096, China
b

h i g h l i g h t s
 A new test was developed to evaluate the permanent deformation of asphalt mixtures.  The ber-reinforced mechanism for asphalt mixture was studied from microscopic view.  We used a varying pressure in PTT to simulate the actual connements in pavement.  A correlation was analyzed between the test results from PTT and wheel tracking test.

a r t i c l e

i n f o

a b s t r a c t
The laboratory tests were conducted to evaluate the effects of polyester ber on permanent deformation of asphalt mixtures. The results indicate the bers improve the deformation resistance of mixtures. The connements in the partial triaxial test (PTT) have signicant effects on the permanent strain of mixtures. The varying connement in the PTT better simulates the actual connements in pavements. A stronger correlation of test results from the PTT and the wheel tracking test was found. It is concluded that the addition of polyester bers improves the mechanical performance of mixtures, and the PTT method is a reliable new test method. 2013 Elsevier Ltd. All rights reserved.

Article history: Received 26 April 2013 Received in revised form 12 September 2013 Accepted 9 December 2013

Keywords: Asphalt mixture Polyester ber Permanent strain Rutting depth Strain rate

1. Introduction Asphalt has been primarily used as a binder to construct asphalt pavements on highway. Asphalt, as a kind of polymeric material, has obvious viscoelastic characteristics [1]. The asphalt pavement is prone to rutting when heavy loads are applied at high ambient temperatures. The accumulation of irrecoverable strain in asphalt pavement layers is one of major causes to cause rutting [2]. Rutting has been one of the common distresses in pavements that affects riding comfort for road users and causes high maintenance costs for road agencies [3]. Modied-asphalt with additives is usually used to change the phase composition and improve the engineering properties of asphalt mixtures. Among various additives, polymer bers are mainly used to enhance the rutting resistance of asphalt pavements [4]. Previous researchers have reported that polymer
Corresponding author. Tel.: +86 25 8542 7747; fax: +86 25 8379 5184.
E-mail address: seuxt@hotmail.com (T. Xu). 0950-0618/$ - see front matter 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2013.12.015

bers not only increased the Marshall stability and reduced the voids in mineral aggregates to improve the cohesion between aggregates [5,6], but also increased the viscosity and stiffness of asphalt mastics and improved asphalt mixtures moisture susceptibility, rutting resistance, fatigue life, and durability [7,8]. In developing an experimental testing method for evaluating the rutting resistance of asphalt mixtures, most researchers have used wheel tracking test, the uniaxial compressive creep test, the triaxial repeated load test (TRT), the indirect tension test, and the bending creep test [9]. The wheel tracking test simulates trafc loading on pavements by applying a wheel load on a slab specimen. The testing conditions are similar to pavements in service and the rut depth is measured after a specic number of loading cycles [3]. The wheel tracking test has been proven to be an effective method to evaluate the rutting potential in asphalt pavements [2]. Another commonly used test to measure the permanent deformation of asphalt mixtures is the TRT, which is conducted by applying repeated vertical loading in a triaxial setup [10].

562

T. Xu et al. / Construction and Building Materials 53 (2014) 561567 Table 2 Properties of the used aggregate. Properties Coarse aggregate angularity (%) Fine aggregate angularity (%) Flat/elongated particles (%) Clay content (%) Coarse aggregate specic gravity (g/cm3) Coarse aggregate absorption (%) Fine aggregate specic gravity (g/cm3) Fine aggregate absorption (%) Sand equivalent (%) Abrasion loss (Los Angeles) (%) Frost action (with Na2SO4) (%) Polishing value Standard ASTM D5821 AASHTO T304 ASTM D4791 AASHTO T 112 ASTM C-127 ASTM C-127 ASTM C-128 ASTM C-128 AASHTO T 176 ASTM DC-131 ASTM C-88 BS-813 Testing results 100 46.8 2.6 0.4 2.843 0.46 2.839 0.63 78 16.7 1.3 0.96

Results obtained from the TRT are typically presented in terms of the cumulative permanent strain versus the number of loading cycles [11]. Goh and You [12] developed a new simple stepwise method to determine the ow number in TRT. Taherkhani [13] investigated the uniaxial and triaxial steady-state deformation behavior of realistic asphaltic mixtures using uniaxial and triaxial creep tests. Although the TRT has been used as an alternative to the wheel-tracking test, there are concerns on the ability of this test as an effective method to distinguish the permanent deformation behavior of different types of asphalt mixtures [2]. Additionally, the constant conning pressure in the TRT was not realistic compared to the eld condition because the connement in the pavement structure changes with temperatures, trafc loads, and the surrounding material. Therefore, the TRT cannot simulate the stress conditions encountered in the real pavements. Moreover, the application of triaxial test, such as the TRT, is limited due to the relatively complicate testing procedure and expensive equipment.

Table 3 Properties of the crushed limestone powder and polyester ber. Material type Properties Particle size Specic gravity CaO content SiO2 content Passing percent in Passing percent in Passing percent in Diameter Length Relative density Melting point Fire point Tensile strength Maximum tension Testing results 00.3 mm 2.773 g/cm3 48.2% 1.46% 99.1% 93.5% 79.5% (0.016 0.0025) mm 6 mm 1.321.40 248 C 536 C 517 34.5 MPa 33 9%

2. Objective In this study, a new creep test was developed, named the partial triaxial test (PTT), to evaluate the permanent deformation behavior of asphalt mixtures. One signicant advantage of the new test is that the connement pressure in the PTT were generated from the surrounding asphalt material and thus changed with temperature, loading, and the mechanical properties of the surrounding material. Thus, the new testing method could better simulate the connement conditions of asphalt mixtures in the led condition than the traditional traiaxial repeated load test (TRT). The focus of this study is to investigate the effects of polyester ber on permanent deformation of asphalt mixtures at the high temperature using the TRT and the PTT. Different loading levels were applied on the specimens to evaluate the development of permanent strain and strain rate under repeated loading cycles. For comparison, the laboratory wheel tracking tests were utilized to directly measure the rutting depth and dynamic stability of the asphalt mixture. In addition, the correlation between the total accumulative rut depth and the strain slope value at the secondary stage in the TRT and the PTT was discussed.

Crushed limestone powder

0.3 mm 0.15 mm 0.075 mm

Polyester ber

strain

3.2. Asphalt mixtures Sup19 and AC20 are often used as the middle layer of a typical 3-layer pavement structure in China. The middle layer is usually subjected to rutting at the high temperature, thus was chosen in this study.

3. Materials and testing method 3.1. Raw materials Polymer (styrenebutadienestyrene (SBS)) modied asphalt binder with a performance grade of PG76-22 was obtained from Zhenjiang Meilun Bitumen Co. Ltd, China. The basic properties of asphalt binder were given in Table 1. The crushed amphibolites aggregate (Pengcheng Aggregate Co., Ltd., China.) was used for preparing asphalt mixtures. The properties of the aggregate are summarized in Table 2. Additionally, the properties of crushed limestone powder and polyester ber provided by manufacturers were shown in Table 3.

3.2.1. Sup19 mixture The preparation of Sup19 asphalt specimens was as follows. Based on the specic gravity of selected aggregates, asphalt binder content was calculated to be 4.1% by the mass of total mixture. The Superpave gyratory compactor (EP-31111model, US) was used to prepare asphalt mixture specimens. The number of gyrations for initial compaction (Ninitial), design compaction (Ndesign), and maximum compaction (Nmax) was 8, 100 and 160 gyrations, respectively. The diameter of the specimen cylinder was 150 mm. The applied axial pressure was 0.6 MPa, and angle of gyration was 1.25, and the revolution of gyratory load was 30 r/min. The loose asphalt mixtures were aged in oven at the compaction temperature for 2 h prior to the compaction. This simulated the aging and absorption of asphalt binder during the construction of hot-mix asphalt pavements [14]. The nal combined aggregate gradation for Sup19 was presented in Fig. 1. The limit values follow the Technical Specication for Construction of Highway Asphalt Pavement (JTG F402004) in China. The optimum asphalt content was found at 4.1 percent.

Table 1 Properties of the SBS modied asphalt binder. Properties Penetration Ductility at 5 C Softening point Viscosity at 135 C Flash point Standard ASTM ASTM ASTM ASTM ASTM D5-61 D113-86 D36-26 D4402 D92 Testing results 5.8 mm 48.7 cm 80.2 C 2.3 Pa s 336 C

3.2.2. AC20 mixture The Marshall method (ASTM D1559) was used to design the AC20 asphalt mixture. The identical cylindrical specimens with the Sup19 mixture (101.6 mm in diameter and 63.5 1.3 mm in height) were produced with 75-blow compaction per side by Marshall Compactor. The nal combined aggregate gradation for AC20 was given in Fig. 1. The optimum asphalt content was found at 4.4 percent. The content of polyester ber was selected 0.3 percent by the total mass of AC20 mixture to evaluate the effect of polyester bers on dynamic creep properties of asphalt mixtures. The optimum asphalt content of the polyester ber-reinforced asphalt mixture was found at 4.5 percent that is greater than the control AC20

T. Xu et al. / Construction and Building Materials 53 (2014) 561567

563

Fig. 1. The combined aggregate gradations for AC20 and Sup19 mixtures.

mixture. This is because the polyester ber may absorb a certain amount of asphalt binder. The other volumetric parameters of the ber-reinforced asphalt mixture meet the AC20 specied requirements.

Fig. 3. Test setup of partial triaxial test (PTT).

3.3. Specimen fabrication 3.3.1. Cylindrical specimen for TRT and PTT Gyratory compaction method was utilized to prepare cylindrical specimens with a diameter of 150 mm and a height of 162 mm (150 mm H162 mm), as shown in Fig. 2(a). The air void content of mixture specimen was controlled at 4.0 percent. To prepare the TRT specimens, the gyratory compacted specimens were rstly cored, and then manufactured to produce the specimen with a diameter of 100 mm and a height of 150 mm (100 mm H150 mm), as shown in Fig. 2(b and c). On the other hand, the specimens used in the PTT were prepared by cutting the gyratory compacted specimens along the middle height (see Fig. 2(a)). In this case, two specimens (150 mm H80 mm) were obtained from one standard gyratory compacted specimen.

3.4. Laboratory testing 3.4.1. Triaxial repeated load test (TRT) In this study, the TRT was conducted to evaluate the dynamic creep properties using three specimens for each type of mixture. The TRT was performed at loading levels of 400, 700 and 1000 kPa with a constant conning pressure of 138 kPa at 60 C [10]. A haversine pulse loading of 0.1 s and an unloading time of 0.9 s were applied until the specimen failed in the tertiary stage or until 10,000 cycles. The number of cycles was continuously recorded during the test.

3.3.2. Slab specimen used in wheel tracking test The rectangular slab specimens of asphalt mixtures were fabricated for wheel tracking test. The slab specimen was compacted using a slab compactor to a certain air voids of 4.0 0.5 percent. The size of the slab samples was 300 mm in length, 300 mm in width, and 50 mm in thickness. Because the air void content is a key factor affecting the total rutting depth in asphalt pavement, different rolling compaction setups were used to ensure the air void contents of the slab specimen in wheel tracking tests are the same as the air void contents of the specimens in the TRT and the PTT. The air void contents of three slab specimens were measured to ensure the good repeatability.

3.4.2. Partial triaxial test (PTT) The testing setup of the new PTT is shown in Fig. 3. The diameter of the top loading platen and base platen were 75 mm, and the diameter of the specimen was 150 mm. The axis of the loading platen, base platen, and the specimen were located in the same vertical line. Two Linear Variable Differential Transformers (LVDTs) were placed on the top surface of the loading platen to record the vertical deformation of the specimen. The wax paper was placed between the loading platen and the top surface of the specimen and between base platen and bottom surface of the specimen. The specimen was put in an air conditioner for at least three hours at a constant temperature before testing. To compare the testing results obtained in PTT and TRT, the same haversine pulse loading was applied on the specimen. The UTM machine was used to apply three levels of loading (400 kPa, 700 kPa and 1000 kPa) at 60 C, the same as the ones used in TRT.

Fig. 2. Schematic diagram of preparation of (a) gyratory compacted specimen, (b) cored specimen and (c) manufactured specimen for TRT.

564

T. Xu et al. / Construction and Building Materials 53 (2014) 561567

Fig. 4. The permanent strain development and strain rate of asphalt mixtures under different load levels in TRT. 3.4.3. Wheel tracking test The test procedure of wheel tracking tests follows the relevant Chinese standard [15]. Rectangular slab specimens of asphalt mixture were put in an air conditioner with a constant temperature of 60 0.5 C for about 6 h. The slabs were placed on a steel plate and conned by steel plates at the boundaries. A loading of 0.7 MPa was applied on the slabs through a rubber wheel with a width of 50 mm. The traveling distance of the wheel was 230 10 mm at a speed of 42 1 cycles/ min for 60 min. LVDTs were used to measure the vertical displacement of the slab specimen. The unrecoverable displacement (rutting depth) was recorded one time per 20 s during test.

in the TRT. The prepared mixtures show signicant difference in dynamic creep property under the same testing conditions. It is expected that the development of permanent strain includes three stages under the three loading levels: primary, secondary and tertiary. In the primary stage, there is a rapid decrease in the strain rate. The secondary stage is characterized by a relatively constant strain rate. The strain rate begins to increase in the tertiary stage that indicates that the specimen begins to deform signicantly and individual aggregate particles move apart from each other [16]. The testing results show that the development of permanent strain only includes two stages under the three loading levels except for AC20 mixture under the loading level of 1000 kPa (see Fig. 4(c)). This suggests that the loading levels of 400 kPa and 700 kPa might be too low for the mixtures in the TRT because few specimens undergo the tertiary stage at the high temperature. From Fig. 4, it was observed that polyester bers have signicant effects on reducing the permanent strain of the mixtures compared to the control mixture (AC20). The ber-reinforced mixture shows the smaller deformation and strain rate, indicating the stronger resistance to deformation [1720]. It is noted that the tertiary stage was not observed in ber-reinforced AC20 mixture even under the loading level of 1000 kPa. This is probably because bers interweave together and form a three-directional network in the asphalt mixture when bers are uniformly dispersed in asphalt mixture, as shown in Fig. 5(a). The addition of bers results in toughening and bridging effects because of the overlap connections between discrete bers. This spatial reinforcing network of polyester bers could reinforce the asphalt mixture and resist damage propagation (see Fig. 5(b)). The three dimensional network of polyester bers could also assist in the formation of a thicker coating of asphalt mastics around the coarse aggregates, as shown in Fig. 5(c). This reduces aggregate shear sliding at the interface and thus the ow of asphalt mixture at the high temperature. The bers could absorb asphalt binder due to their larger specic surface areas, leading to an increase in the viscosity and stiffness of asphalt mastics. This increase is benecial to improve the interface adhesion between asphalt binder and ber (see Fig. 5(d)). The increase in the viscosity and stiffness of asphalt mastics will contributes to the resistance to deformation or rutting at the high temperature. On the other hand, the AC20 mixture exhibits a higher permanent strain and strain rate compared to the Sup19 mixture. This is attributed to the weaker aggregate skeleton and the higher asphalt content in the AC20 mixture. A stronger aggregate skeleton is formed in Sup19 mixture than that in AC20 mixture. The stability of aggregate skeleton is improved by increasing the number of contact points among coarse aggregates. The stress is transferred through the contact points in the aggregate skeleton. The greater internal friction among Sup19 aggregates is benecial to bear the load, increasing the high-temperature stability of Sup19 mixture. Moreover, the redundant asphalt in the AC20 mixture plays a role of lubricant at high temperature, decreasing the internal friction among aggregates. Therefore, AC20 mixture shows a higher permanent strain and strain rate. 4.2. PTT test results The PTT test results of asphalt mixtures are shown in Fig. 6. Similar to the strain development in the TRT, the permanent strain is divided into primary, secondary and possible tertiary stages depending on the applied stress level. From Fig. 6, it was found that the ber-reinforced asphalt mixture showed good resistance to permanent deformation. This is attributed to the three-dimensional spatial networking effect of bers dispersed in the asphalt mixture and the effects of bers on asphalt absorption and stabilization, as discussed above.

4. Results and discussion 4.1. TRT results Fig. 4 illustrated the average axial permanent strain and strain rate of asphalt mixtures with the number of loading repetitions

T. Xu et al. / Construction and Building Materials 53 (2014) 561567

565

Fig. 5. SEM image of (a) bers; (b) toughening and bridging effects of bers in asphalt mixture; (c) reinforced asphalt mastics around the coarse aggregates and (d) bonding of bers and asphalt mastics matrix.

Interestingly, the measured permanent strain in the PTT was found greater than the strain measured in the TRT at the same loading cycles. In addition, a greater maximum strain rate and a more rapid decrease in the strain rate were observed under the rst hundreds of loading cycles in the PTT. The reason for this is that the specimen in the PTT provides a limited lateral connement when the load is only applied on the center area of the specimen surface and this connement keeps changing during the test. When micro-damage appears in the asphalt mixture specimen, the connement is weakened by these micro damages. In contrast, the connement in the TRT keeps constant during the test because the conning pressure is applied by the testing equipment. Hence, the connement in the TRT would be greater than the connement in the PTT. Therefore the PTT better simulates the conned conditions in the actual pavement compared to TRT.

Figs. 7 and 8 show that the AC20 asphalt mixture exhibits the greater rutting depth and lower deformation rate and dynamic stability compared to the Sup19 mixture. The results further validate that the polyester ber can improve the rutting resistance of asphalt mixtures. The reasons for this are a weaker aggregate skeleton is formed and the higher asphalt content in AC20 mixture than that in Sup19 mixture, decreasing the internal friction among AC20 aggregates. Also, the toughening and bridging effects of polyester bers reinforce the asphalt mixture and resist deformation. 4.4. Relationship of results between wheel-tracking test and TRT and PTT In the secondary stage of permanent strain development in the TRT and the PTT, the permanent strain accumulates per cycle is relatively constant with loading cycles. This allows characterizing the rutting resistance of asphalt mixtures using a single parameter related to the rate of rutting. It is assumed here that the relationship between the permanent strain and the loading cycles can be described using a linear model as shown in Eq. (2) [21].

4.3. Wheel-tracking test results Two parameters were used to assess the performance of asphalt mixtures during wheel tracking test. The rut depth was measured as the primary measure of the rutting resistance and the deformation rate was used as a secondary measure. The measured rutting depths and deformation rates of asphalt mixtures are presented in Fig. 7. The total accumulative rutting depth and dynamic stability of asphalt mixtures are shown in Fig. 8. The dynamic stability (DS) was dened in Eq. (1).

ep eFs C N F s

DS

N 15 42 15 d60 d45 d60 d45

where N is the number of wheel loading cycles per minute, (N = 42 cycles/min in this study), and d60 and d45 are the rutting depths measured at 45 and 60 min, respectively.

where ep is the permanent strain; N is the number of total loading cycles; C is the slope value of permanent strain in the secondary stage; Fs is the number of loading cycle at the starting point of the secondary stage; eF s aF b s is the permanent strain at the starting point of the secondary stage; a and b are constants determined using tests results. To evaluate the relationship of testing results between the wheel-tracking test and the TRT and the PTT, the correlation between the total accumulative rutting depth and the slope (C) of permanent strain at the secondary stage in the TRT and the PTT was analyzed. As shown in Figs. 9 and 10, a linear relationship was found with high R-square values.

566

T. Xu et al. / Construction and Building Materials 53 (2014) 561567

Fig. 7. The rutting depth development and deformation rate of different asphalt mixtures.

Fig. 8. The total accumulative rutting depth and dynamic stability of different asphalt mixtures.

Fig. 6. The permanent strain development and strain rate of asphalt mixtures under different load levels in PTT.

Fig. 9. The correlation of the rutting depth and slope value in TRT.

Although linear relationship were found between the rutting depths and the strain slopes under three loading levels, a stronger correlation was observed at the loading level of 700 kPa than the other two loading levels. This is because the same loading level of 700 kPa was used in the wheel tracking test. Additionally, the strain slope values from the PTT shows a greater correlation with the rutting depth than the strain slope values from the TRT. This is because the conning stress state of the asphalt mixture in the PTT is closer to the test condition in the wheel tracking test. This further indicates that the PTT is a more reliable test to evaluate the permanent deformation or rutting resistance of asphalt mixtures.

Fig. 10. The correlation of the rutting depth and slope value in PTT.

T. Xu et al. / Construction and Building Materials 53 (2014) 561567

567

5. Conclusions and further recommendations The main conclusions are summarized as following: (1) The polyester bers improve the deformation resistance of asphalt mixtures because of networking effects, reinforcement, asphalt absorption and stabilization. The application of polyester bers changed the characteristics of asphalt mixture in a benecial way. (2) Lateral connements have effects on the permanent strain of asphalt mixtures. The varying connement in the PTT better simulates the actual conning conditions of asphalt pavement compared to the TRT. Additionally the permanent strain measured in the PTT was greater than the one measured in the TRT at the same loading level. (3) A strong correlation was found between the strain slope value at the second stage in the PTT and the total rut depth in the wheel tracking test. This indicates that the PTT is more reliable to evaluate permanent deformation of asphalt mixtures compared to the traditional TRT. In the future, the relationship of testing results between in-place performance of asphalt pavement and TRT and PTT of asphalt mixture are needed to further validated. Acknowledgements This study is funded by the National Natural Science Foundation of China (NSFC, Grant No. 51378264) and the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). References
[1] Yang Q, Ning J. The environmental inuence of asphalt pavement and countermeasures. Energy Proced 2011;5:24326.

[2] Aksoy A, Iskender E. Creep in conventional and modied asphalt mixtures. Proc Inst Civil Eng-Transp 2008;4:18595. [3] Ahmad J, Abdul RMY, Hainin MR. Rutting evaluation of dense graded hot mix asphalt mixture. Int J Eng Technol 2011;5:5660. [4] Chen H, Xu Q. Experimental study of bers in stabilizing and reinforcing asphalt binder. Fuel 2010;89:161622. [5] Wu S, Ye Q, Li N. Investigation of rheological and fatigue properties of AC mixtures containing polyester bers. Constr Build Mater 2008;22:21115. [6] Chen H, Xu Q, Chen S, Zhang Z. Evaluation and design of ber-reinforced asphalt concrete mixtures. Mater Des 2009;30:2595603. [7] Putman BJ, Amirkhanian SN. Utilization of waste bers in stone matrix AC mixtures. Resour Conserv Recycl 2004;42:26574. [8] Ye Q, Wu SP, Li N. Investigation of the dynamic and fatigue properties of ber modied asphalt mixtures. Int J Fatigue 2009;31:1598602. [9] Qi X, Witczak MW. Time-dependent permanent deformation models for asphaltic mixtures. Res Record 1998;1639:8393. [10] Witczak MW, Kaloush K. Simple performance test for superpave mix design. (NCHRP Report-465). Washington [DC]: National Academy Press; 2002. [11] Bhasin A, Button JW, Chowdhury A. Evaluation of simple performance tests on hot-mix asphalt mixtures from south central United States. Transp Res Record 2004;1891:17481. [12] Goh SW, You ZP. A simple stepwise method to determine and evaluate the initiation of tertiary ow for asphalt mixtures under dynamic creep test. Constr Build Mater 2009;23:3398405. [13] Taherkhani H. Compressive creep behaviour of asphalt mixtures. Proced Eng 2011;10:5838. [14] Prowell BD, Zhang J, Brown R. Aggregate properties and the performance of Superpave designed hot mix asphalt. (NCHRP Report-539). Washington [DC]: National Academy Press; 2005. [15] Li FP, Yan EH, Huang SC. Standard test methods of bitumen and bituminous mixtures for highway engineering. Beijing: China Communications Press; 2011. [16] Wang H, Al-Qadi IL, Faheem AF, Bahia HU. Effect of mineral ller characteristics on asphalt mastic and mixture rutting potential. J Trans Res Board 2011;2008:339. [17] Laureano RH, Anand JP, Carlos A. Characterization of cement-ber-treated reclaimed asphalt pavement aggregates: preliminary investigation. J Mater Civ Eng 2011;23:97789. [18] Tapkin S. The effect of polypropylene bers on asphalt performance. Build Environ 2008;43:106571. [19] Lee SJ, Rust JP, Hamouda H, Kim YR, Borden RH. Fatigue cracking resistance of ber-reinforced asphalt concrete. Text Res J 2005;75:1238. [20] Abtahi SM, Sheikhzadeh M, Hejazi SM. Fiber-reinforced asphalt-concrete a review. Constr Build Mater 2010;24:8717. [21] Zhou F, Scullion T, Sun L. Verication and modeling of three-stage permanent deformation behavior of asphalt mixes. J Trans Eng 2004;130:48693.