Construction and Building Materials 136 (2017) 184–191

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Fatigue performance evaluation of modified asphalt binder using ofdissipated energy approach

http://dx.doi.org/10.1016/j.conbuildmat.2017.01.0100950-0618/� 2017 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Department of Chemistry, Faculty of Science, IslamicAzad University, East Tehran Branch, Tehran, Iran.

E-mail address: massoumeh.abbasi@yahoo.com (M. Abbasi).1 Professor.2 Ph.D. Research Scholar.3 Assistant Professor.4 Ph.D. Research Scholar.5 Assistant Professor.

Mahmoud Ameri a,1, Mohammadreza Seif b,2, Massoumeh Abbasi c,3,⇑, Mohammad Molayemd,4,Alireza KhavandiKhiavi e,5

a School of Civil Engineering, Iran University of Science and Technology, Narmak, Tehran, 16846-13114, Iranb School of Civil Engineering, Iran University of Science and Technology, Tehran, Iranc Islamic Azad University, East Tehran Branch, Department of Chemistry, Faculty of Science, Tehran, PO Box 33955-163, Irand School of Civil Engineering, Iran University of Science and Technology, Narmak, Tehran, IraneDepartment of Civil Engineering, University of Zanjan, Zanjan, Iran

h i g h l i g h t s

� The modified asphalt binders perform better at high strain levels.� The initial dissipated energy has the potential to become a rapid method to describe the fatigue resistance of asphalt.� In nonlinear strains, the curve slope of modified binder is less than curve slope of pure binder.� In nonlinear strains, unmodified and modified asphalt binders with increasing loading cycles, DE decreased.

a r t i c l e i n f o

Article history:Received 7 August 2016Received in revised form 16 December 2016Accepted 3 January 2017Available online 19 January 2017

Keywords:Asphalt binderFatigue lifeTime sweep testDissipated energyIDE

a b s t r a c t

In this paper dissipated energy (DE) versus loading cycles is plotted, at the same time fatigue life ofasphalt binders were investigated using of dissipated energy (DE) approach and based on linear and non-linear strain levels. It was found that based on results obtained of time sweep test; there is a fair corre-lation between dissipated energy at the 50th loading cycle and fatigue life of asphalt binders in all strainlevels. As a result, initial dissipated energy (IDE) is a parameter obtained from time sweep test, so theamount of initial dissipated energy (IDE) at the 50th loading cycle can be predicted the fatigue of asphaltbinders without performing millions of loading cycles.

� 2017 Elsevier Ltd. All rights reserved.

1. Introduction

Bituminous materials are widely used in construction of flexiblepavements. Fatigue cracking is one of the major distresses of flex-ible pavements and is defined as the damage in asphalt pavementsby repetitive stresses and strains due to traffic loading and envi-ronmental factors [1,2,3]. According to report NCHRP-RPT-459,there are several factors effective in estimating and prediction

fatigue life asphalr binders [4,5]. The six most important factorsthat influence on fatigue life asphalr binders are binder type, strainlevel, temperature, frequency, number of cycles and rest period. Ithas been shown that the asphalt mixture fatigue resistance isstrongly correlated with asphalt binder fatigue properties at inter-mediate temperatures [6,7,9,9].

Utilizing the concept of energy in assessing the fatigue behaviorof asphalt binders has two advantages. First, the energy is a scalarfunction which is independent of the direction of the applied stressor strain. Secondly, it includes stress, strain the material’s proper-ties in a single unit of measure [22]. Evaluation of binder fatigueperformance under cyclic loading has continuously been one ofthe most popular research topics for engineers during the pastfew years [10].

Airey et al. [8,11] conducted stress sweep tests using of DSR toobtain the linearity limits of various asphalt binders at different

M. Ameri et al. / Construction and Building Materials 136 (2017) 184–191 185

temperatures. Furthermore, Bahia in 1999 reported the fatigueperformance of asphalt binders at various strain levels in the linearand nonlinear range and their effects on mixture performance[4,5].

There are several developments in terms of asphalt binder fati-gue tests which time sweep test, according to project NCHRP 9-10,has reported to be an effective test method to evaluate binder fati-gue life [4,6,7,12]. So that Time sweep (TS) test is reported to be apromising method to estimate fatigue performance of asphalt bin-ders. This test provides a simple method of applying repeatedcycling of stress or strain loading at selected temperatures andloading frequencies. Another advantage of this test is that it canbe used to calculate fatigue life of asphalt binders based on dissi-pated energy approach. Dissipated energy approach has beenbroadly used as a simplified method to characterize fatigue perfor-mance of viscoelastic materials based on modulus and phase anglein each loading cycles. Van Dijk introduced the approach to evalu-ate fatigue performance of viscoelastic materials in terms of dissi-pated energy [13,14]. He proposed equation (1) to calculatedissipated energy during a single load cycle.

Wn ¼ prnen sin hn ð1Þ

whereWn n is the dissipated energy, rn and en are the applied stressand strain, hn is the phase angle and n represents number of loadingcycles. In this research study in order to have a more meaningful ofthe fatigue properties of asphalt binders, the dissipated energy

Table 1The properties of neat binders used in this study.

Parameters Unit

Specific gravity g/cm3

Penetration (100 g) 0.1 mmSoftening point �CDuctility CmSolubility (%)Flash point (Cleveland) �CKinematic viscosity 135 �C CentistokesHeating loss (%)Penetration after heating loss 0.1 mmPenetration after heating loss original penetration (%)Ductility after heating loss cmPenetration Index (PI) –PVN(25-135) –

Table 2Properties of SBS and CR used in this study.

Physical properties Unit Spec

Density g/cm3 0.94Volatile material wt% MaxStyrene content % 31

Mechanical PropertiesMelt index g/10 min <1Hardness T Shore A 79T.S.V. cst 13.4OtherAppearance – PelleStructure Melt Index: 200 C, 5 kg – Line

Crumb rubberSpecific gravity (gcm3)Moisture content (wt%)Ash content (wt%)Carbon black content (wt%)Extract content (acetone and chlo(wt%)Sulphur content * (wt%)Source

approach was utilized. According to this concept, the dissipatedenergy in each cycle can be determined by Eq. (1).

Cheng et al. (2002) showed that the energy balance is influ-enced by rheological properties of the mix and the asphalt binder,and also fatigue damage in viscoelastic materials can be due tostored and dissipated energies [15]. Eq. (2) was proposed to char-acterize the number of cycles to failure using dissipated energy.Where: Nf is number of load application to failure, Wi is dissipatedenergy and K1, K2 are experimentally determined coefficients.

Nf ¼ K11Wi

� �K2

ð2Þ

Baburamani and Porter [16], Rowe [17] and Ghuzlan [18] indi-cated that initial dissipated energy also can be an acceleratemethod to evaluate fatigue life of viscoelastic materials[16,17,18]. Initial dissipated energy (IDE) is defined as the dissi-pated energy measured at 50th loading cycle. However, Carpenterand Shen (2003) found that the initial dissipated energy approachis not appropriate for the whole loading range, especially at lowstrain fatigue tests [19].

The objective of this study is to determine the linear and nonlin-ear strain ranges of six asphalt binders in a constant frequencyusing of strain sweep test. In during to strain sweep tests, for deter-mining of linear and nonlinear range, strain 0–20% applied on allasphalt binder samples. Next, dissipated energy versus loadingcycles is plotted on linear and non linear strain levels. According

Test methods 60/70 85/100 Tabriz

ASTM AASHTO

D70 T228 1.016 1.009D5 T49 66 99D36 T53 51 46.4D113 T51 >100 >100D2042 T44 99.5 99.56D2170 T201 303 260D6 – 361 240– – 0.04 0.15– – 47 66– – 85.4 66– – >50 cm 80– – �0.73 �0.3– – �0.39 3.21

ification Test method

ISO 2781. 0.5 ASTM D1416

LSYQS1DO11

ASTM D1238ASTM D2240ASTM D445

t –ar –

ACR CCR1.042 1.0530.76 0.776.01 4.6632.98 30.41

roform) 9.86 11.69

2.02 1.24Car tires Car tires

186 M. Ameri et al. / Construction and Building Materials 136 (2017) 184–191

to results the correlation between DE, IDE and fatigue life of sixasphalt binder types were computed using of time sweep tests atfive strain levels of 1, 2, 3, 4 and 6 percent. The frequency and load-ing temperature were 10 Hz and 25 �C respectively.

2. Materials and experiments

2.1. Basic materials

In this research work, 60–70 penetration grade asphalt (PG 61-16) and 85–100 penetration grade asphalt (PG 58-22) providedfrom Pasargad Oil Company were selected as neat asphalt binders.Two kinds of additives were used as modifier in the 85–100 asphaltbinder. Modifiers were the crumb rubber (CR) and Styrene-Butadiene-Styrene (SBS) polymer. The CR particles were added at8 and 16 percent by the weight of the neat bitumen. The corre-sponding amounts of SBS added to 85–100 asphalt binder were 3and 5 percent by the weight of the neat bitumen. Mixing procedurewas carried out by means of a high shear mixing for 30 min at thespeed of 4000 rpm. According to previous researches, �40 meshrubber enhances was employed to produce rubber modified bin-ders [20]. The mixing temperature was 180 �C. Properties of neatasphalt binders used in this study and modifiers are listed in Tables1 and 2.

2.2. Testing program

2.2.1. Strain sweep testIn this research study all asphalt binders were aged through the

rolling thin film oven (RTFO) test according to ASTM D2872-12Standard (D2872-88, 1995) and also the suggested method byBahia et al. was applied [21]. For reliability, at least two replicationof RTFO was conducted on each sample.

Strain sweep test is conducted using a dynamic shear rheome-ter (DSR) to determine linear viscoelastic and nonlinear viscoelas-tic limits of unmodified and modified asphalt binders. DSR tests

Fig. 1. Result strain sweep test for unmo

were conducted according with NCHRP report 459 at constantstrain mode of loading under the frequency of 10 Hz at the temper-ature of 25 �C. This situation was employed in all tests to obtain thelinear viscoelastic range of asphalt binders. In strain sweep test, thestrain level was increased from 1% to 20% and the dynamic shearmodulus (G⁄) was recorded. According to Anderson et al. [2], theliner visco-elastic limit of materials is defined as the point wherecomplex modulus, (G⁄) decreased to 95% of its initial value [2].

2.2.2. Time sweep testTime sweep (TS) test is reported to be a promising method to

estimate fatigue performance of asphalt binders. So that, six differ-ent binders were prepared and their dissipated energy at differentloading cycles by time sweep test was investigated and correlatedwith the fatigue life criterion.

Time Sweep (TS) test was introduced during NCHRP 9-10 fordetermine asphalt binder fatigue [5,10,24]. As already mentionedthe time-sweep test measured the fatigue life and dissipatedenergy of asphalt binders using of DSR apparatus with an 8 mmplate under controlled-strain mode and at a constant frequencyof 10 Hz. All samples including unmodified and modified asphaltbinders were aged through the Rolling Thin Film Oven (RTFO) testto represent the short term aging of asphalt binder [21,23]. Thefatigue failure criterion is selected as a 50 percent reduction ininitial complex shear modulus (G⁄). In order to have a more mean-ingful of the fatigue properties of asphalt binders, the dissipatedenergy approach was applied. According to this concept, thedissipated energy is also calculated for each loading cycle usingequation (3). The test was conducted at 5 different strain levelsranging from 1% to 6% with increment of 1%.

DEi ¼ pe2Gi sin dI ð3ÞIn where: DEi is dissipated energy at cycle i, e is strain level and

di is phase angle at cycle i.The total approximately 175000 data were analyzed and

compared.

dified and modified asphalt binders.

M. Ameri et al. / Construction and Building Materials 136 (2017) 184–191 187

3. Results and discussion

3.1. Strain sweep test

The strain sweep test was developed to identify the linear vis-coelastic and nonlinear viscoelastic response of asphalt binders.

Fig. 2. The variations of dissipated energy (DE) per load

Fig. 3. The mean, maximum and minimum of DE for all of t

So the response to asphalt binder can be divided into two parts lin-ear and nonlinear based on the applied strain. According to ourobservations, linear viscoelastic performance all type of asphaltbinders used in this research study was between 0 and 1 percentof strain level. The result of strain sweep tests is shown in Fig. 1.It was observed that neat 60/70 asphalt binder has the highest

ing cycle (LC) in linear and nonlinear strain levels.

he asphalt binders in linear and nonlinear strain levels.

188 M. Ameri et al. / Construction and Building Materials 136 (2017) 184–191

complex modulus in low strain levels (1%). However, in high strainlevels the complex modulus of neat 60/70 asphalt binder and 3%SBS are almost equal, in which represents the modified binder per-formed better in high strain levels. The same trend is seen for 8%rubber and the neat 85/100 asphalt binder.

3.2. Time sweep test

According to Fig. 2, dissipated energy plotted versus loadingcycles for six binders employed in this study on linear strain 1%and nonlinear strain levels (2, 3, 4, 6%) in a logarithmic scale. Itwas found that in the nonlinear strain levels, regression of dissi-pated energy (DE), in terms of loading cycles and based on strainlevel follows of linear equations. However, it is seen that DE atthe strain level of 1% where corresponds to linear region of theasphalt binders, there is no linear regression. In addition, it wasfound that in nonlinear strains (2%, 3%, 4% and 6%) the variationDE varsue loading cycles followed from linear regression after1000 cycles whereas in linear stain (1%) same trend obtained after10000 cycles. Hence, according to Fig. 2, at the nonlinear strain

Fig. 4. The variation of mean, standard deviation, maximum and m

Fig. 5. The dissipated energy (DE) in the cycle of failure and the initial dissipated energy (

levels, the results show that the curve slope in modified asphaltbinders is less than curve slope of unmodified asphalt binder. Inother words, in a special cycle, variation of DE in modified asphaltbinder is less than unmodified asphalt binder whereas the total ofdissipated energies in the modified asphalt binders are more thanmodified ones.

In Fig. 3, based on the asphalt binder type and strain level, acomparison of various types of DE as mean, maximum (max) andminimum (min) is provided. It is observed that in nonlinear region,there is a similar trend based on binder type and strain level. How-ever, in linear range there is not any specific trend. In addition,according to Fig. 3, with increasing of loading cycles, DE isdecreased in nonlinear strain levels for both unmodified and mod-ified asphalt binders. Furthermore, among of modified binders,rubber powder showed lowest mean, maximum and minimum ofDE in comparing to other samples.

Moreover, it was seen that 5% SBS modified 85–100 asphalt bin-der has the highest fatigue life and neat 60–70 asphalt binder hasthe lowest fatigue life among all samples. It can also be inferredfrom the results that the fatigue life of all asphalt binders

inimum of Nf in linear and nonlinear strains for binder types.

IDE) of binder types during time sweep test in comparison with binder fatigue lives.

Fig. 6. (a) The correlation between dissipated energy (DE) at the cycle of failure and asphalt binder fatigue life (Nf). (b) The correlation between initial dissipated energy (IDE)and the fatigue life (Nf) of asphalt binders.

Fig. 7. The correlation between applied loading strain and fatigue life of asphaltbinders.

M. Ameri et al. / Construction and Building Materials 136 (2017) 184–191 189

decreased drastically when the strain level changes from 1% to 2%.This performance is due to the change in the viscoelasticity of allasphalt binders from linear to nonlinear. As a result the fatiguelives of all asphalt binder decrease with increasing the strain level.

In Fig. 4, the variation of mean, standard deviation, maximumand minimum of Nf in linear and nonlinear strain levels fordifferent types of asphalt binders is plutted. According to Fig. 4,in the pure binders (60–70 and 85–100) statistical fluctuations(maximum, minimum, mean and standard deviation) of Nf is lessthan modified binder and SBS 5% showed a maximum value (orpeak). In addition, modified binder with rubber powder displayeda fewer fluctuations. Also in the strain 2% and higher does have lessfluctuations and these fluctuations are much greater in strain 1%.

As mentioned previously, the dissipated energy concept wasused to determine the fatigue life of modified and unmodifiedasphalt binders. Fig. 5 shows the correlation of dissipated energy(DE) and initial dissipated energy (IDE) with the fatigue life ofasphalt binders in different strain levels during time sweep test.Two fatigue failure criterion is selected in which the first criterionis a 50% reduction in initial complex shear modulus (G⁄) and lattercriterion is dissipated energy in the 50th cycle of loading. Accord-ing to Fig. 5, there is a fair correlation between the dissipatedenergy and initial dissipated energy values and binder fatigue life.

The correlation between the initial dissipated energy and fati-gue life of asphalt binders is shown in Fig. 6(a), and the correlationbetween dissipated energy in the cycle of failure and fatigue life ofasphalt binders in is shown in Fig. 6b. It can be inferred that initialdissipated energy correlates well with the fatigue life of asphaltbinders and has the potential to describe the ability of asphalt bin-ders to resist fatigue cracking. Dissipated energy in the cycle of fail-ure has also a strong correlation with the fatigue life of asphaltbinders.

The correlation between the fatigue life of asphalt binders andloading strain (traditional) was also investigated and is illustratedin Fig. 7. As it is shown, the correlation between loading strain andfatigue life of asphalt binders is weak but the initial dissipatedenergy has a stronger correlation with fatigue life of asphaltbinders.

The equation of linear regressions fitted between three vari-ables (DE, IDE, strain level) and fatigue life of asphalt bindersare also investigated. The coefficients, error and R square valuesof the linear regression fitted with fatigue life of asphalt binders

and variables of DE, IDE and strain level are listed in Table 3.As it is shown, the initial dissipated energy and the dissipatedenergy in cycle of failure have a strong correlation with binderfatigue life. In addition to having a strong correlation with thebinder fatigue life, initial dissipated energy is a time savingmethod to predict fatigue life of asphalt binders when comparedto time sweep test.

The time needed to complete time sweep test of asphalt bindersat different strain levels is depicted in Fig. 8. As shown in Fig. 8 thetime sweep test needs an enormous amount of time to be com-pleted especially when modified binders are used. Whereas, initialdissipated energy needs only 5 s to be determined.

Table 3Linear regression coefficients fitted between binder fatigue life and three variables of DE, IDE and strain level. B is the slope of the fitted line and C is the regression intercept.

Dependent Variable: log.Nf

Independent Variable: logDE Independent Variable: logIDE Independent Variable: logЄ

Binder Model Summary Parameter Estimates Model Summary Parameter Estimates Model Summary Parameter Estimates

RSquare

Std. Errorof the Estimate

C(Constant)

B RSquare

Std. Errorof the Estimate

C(Constant)

B RSquare

Std. Errorof the Estimate

C(Constant)

B

60–70 0.976 0.150 9.542 �1.538 0.982 129 9.788 �1.494 0.971 0.165 �0.525 �2.77885–100 0.999 0.027 8.794 �1.350 0.998 0.037 9.153 �1.342 0.997 0.047 �0.182 �2.609RUBBER 8% 0.984 0.138 10.027 �1.731 0.984 0.138 10.690 �1.773 0.974 0.176 �0.828 �3.151RUBBER 16% 0.998 0.054 11.041 �1.906 0.992 0.107 11.556 �1.902 0.999 0.044 �1.060 �3.567SBS 5% 0.962 0.242 11.320 �1.893 0.962 0.241 12.106 �1.949 0.970 0.215 �0.795 �3.550SBS 3% 0.995 0.079 9.244 �1.403 0.996 0.066 9.810 �1.444 0.996 0.072 �0.876 �3.236ALL 0.904 0.292 10.023 �1.639 0.899 0.173 10.537 �1.650 0.843 0.374 �0.711 �3.149

Fig. 8. Maximum time of completing time sweep test for each binder type.

190 M. Ameri et al. / Construction and Building Materials 136 (2017) 184–191

4. Conclusions

In this research, the potential of dissipated energy and initialdissipated energy to describe fatigue resistance of asphalt binderswere investigated. For this purpose, at first a strain sweep test wasconducted to determine the viscoelastic behavior of asphalt bin-ders. Then, the time sweep test was conducted on 6 different bin-ders in 5 different strain levels and a total number of 175817 datawere analyzed. The conclusions are as follows:

� From the strain sweep test, the linear viscoelastic range ofasphalt binders used in this study is 0 to 1%.

� It was found that in nonlinear strains 2%, 3%, 4% and 6% the vari-ation DE varsue loading cycles followed from linear regressionafter 1000 cycles but in linear stain 1% same trend obtainedafter 10000 cycles.

� In presence SBS 5% (85–100) and in nonlinear strains, mean ofDE reduced and there is a further decline for rubber powder8% and 16%. In contrast, in linear strain 1%, and in presence ofSBS 5%, mean of DE increased.

� In pure binder, statistical fluctuations of Nf (maximum, mini-mum, mean and standard deviation) is less than modified bin-der and these in presence of SBS 5% show a maximum valueor peak. Modified binders with rubber powder display a fewerfluctuations. Also in the strain 2% and higher (nonlinear), these

fluctuations are less and in strain 1% fluctuations are muchgreater.

� In nonlinear strains, the curve slope of modified binder is lessthan curve slope of pure binder. In other words, in a specialcycle, variation of DE in modified binder is less than pure bin-der, but the total of dissipated energy in the modified bindersis more than pure binders.

� In nonlinear strains, unmodified and modified asphalt binderswith increasing loading cycles, DE decreased, but in linearstrain, in beginning of loading cycles, DE increased and thendecreased.

� The modified asphalt binders perform better at high strainlevels. Some modified binders have lower complex modulusthan neat asphalt binders in low strain level whereas, highercomplex modulus at higher strain levels.

� The fatigue life of asphalt binders decreases dramatically whenthey are subjected under strain levels on which their perfor-mance is nonlinear.

� The initial dissipated energy has the potential to become a rapidmethod to describe the fatigue resistance of asphalt binders asit showed a fair correlation to the fatigue life of asphalt binders.

Acknowledgment

The authors gratefully acknowledge the funding support byIranian Gas Engineering and development Co. http://www.nigceng.ir/Pages/home.aspx.

References

[1] C.F. Quintero, Q.L. Momm, L.F.M. Leite, L.L.B. Bernucci, Effect of asphalt binderhardness and temperature on fatigue life and complex modulus of hot mixes,Constr. Build. Mater. 114 (2016) 755–762.

[2] D.A. Anderson, D.W. Christensen, H. Bahia, R. Dongre, M. Sharma, C.E. Antle,Binder characterization and evaluation, volume 3: Physical characterization,Strategic Highway Research Program, National Research Council, Report No.SHRP-A-369, (1994).

[3] H. Ziari, R. Babagoli, M. Ameri, A. Akbari, Evaluation of fatigue behavior of hotmix asphalt mixtures prepared by bentonite modified bitumen, Constr. Build.Mater. 68 (2014) 685–691.

[4] H.U. Bahia, D. Hanson, M. Zeng, H. Zhai, M. Khatri, M. Anderson, A ProjectNchrp 9–10 Superpave Protocols For Modified Asphalt Binders. Draft TopicalReport (Task 9), Prepared For National Cooperative Highway ResearchProgram, Transportation Research Board, National Research Council, (2000).

[5] H.U. Bahia, H. Zhai, K. Bonnett, (1999). Journal of the Association.asphaltrubber.org http://www.asphaltrubber.org/ARTIC/Reports/RPA_A1683.pdf.

[6] D.A. Anderson, Y.M.L. Hir, M.O. Marasteanu, J.-P. Planche, D. Martin, G.Gauthier, Evaluation of fatigue criteria for asphalt binders, Transp. Res. Rec.1766 (2001) 48–56.

[7] D. Anderson, Y. Hir, M. Marasteanu, J.-P. Planche, D. Martin, G. Gauthier,Evaluation of fatigue criteria for asphalt binders, Transp. Res. Record: J. Transp.Res. Board (2001) 48–56.

[8] G. Airey, B. Rahimzadeh, A. Collop, Linear viscoelastic performance of asphalticmaterials, Road Mater. Pavement Des. 4 (3) (2003) 269–292.

M. Ameri et al. / Construction and Building Materials 136 (2017) 184–191 191

[9] M. Ameri, S. Nowbakht, M. Molayem, M.H. Mirabimoghaddam Constructionand Building Materials. A study on fatigue modeling of hot mix asphaltmixtures based on the viscoelastic continuum damage properties of asphaltbinder, 106 (2016) 243–252.

[10] C. Wang, H. Zhang, C. Castorena, J. Zhang, Y.R. Kim, Identifying fatigue failure inasphalt binder time sweep tests, Constr. Build. Mater. 121 (2016) 535–546.

[11] G.D. Airey, B. Rahimzadeh, Combined bituminous binder and mixture linearrheological properties, Constr. Build. Mater. 18 (2004) 535–548.

[12] A.H. Al-Haddad, Fatigue evaluation of Iraqi asphalt binders based on thedissipated energy and viscoelastic continuum damage (VECD) approaches, J.Civ. Eng. Constr. Technol. 6 (2015) 27–50, http://dx.doi.org/10.5897/JCECT2015.0356.

[13] W. Van Dijk, Practical fatigue characterization of bituminous mixes, J. Assoc.Asphalt Paving Technol. 44 (1975) 38–74.

[14] W. Van Dijk, W. Vesser, The energy approach to fatigue for pavement design, J.Assoc. Asphalt Paving Technol. 46 (1977) 1–40.

[15] D. Cheng, Surface free energy of asphalt-aggregate system and performanceanalysis of asphalt concrete based on surface energy, Ph.D, Texas University,(2002).

[16] P. Baburamani, D. Potter, dissipated energy approach to fatigue characteristicsof asphalt mixes, in: Combined 18th ARRB Transport Research Conference andTransit New Zealand Land Transport Symposium, 2–6 September 1996,Christchurch, New Zealand Part 2, (1996).

[17] G. Rowe, Performance of asphalt mixtures in the trapezoidal fatigue test,Asphalt Paving Technol. 62 (1993). 344-344.

[18] K.A. Ghuzlan, Fatigue damage analysis in asphalt concrete mixtures basedupon dissipated energy concepts, University of Illinois at Urbana-Champaign,(2001).

[19] S. Carpenter, K. Ghuzlan, S. Shen, Fatigue endurance limit for highway andairport pavements, Transp. Res. Rec.: J. Transp. Res. Board (2003) 131–138.

[20] A. Ameli, R. Babagoli, M. Aghapour, Laboratory evaluation of the effect ofreclaimed asphalt pavement on rutting performance of rubberized asphaltmixtures, Pet. Sci. Technol. 34 (2016) 449–453.

[21] H.U. Bahia, H. Zhai, A. Rangel, Evaluation of stability, nature of modifier, andshort-term aging of modified binders using new tests LAST, PAT, and modifiedRTFO, Transp. Res. Rec. 1638 (1998) 64–71.

[22] Running, D.L., 1996. An Energy-Based Model for Soil Liquefaction. Ph.D. thesis.Department of Civil and Environmental Engineering, Washington StateUniversity, pp. 77.

[23] AASHTO T 240–09, 2009. Standard Method of Test for Effect of Heat and Air ona Moving Film of Asphalt. American Association of State and HighwayTransportation Officials.

[24] Bahia, H.U., Hanson, D., Zeng, M., Zhai, H., Khatri, M.A.. Anderson, R.M., 2001.Characterization of modified asphalt binders in Superpave mix design. NCHRPReport 459. National Academy Press, Washington (DC).