Materials and Design 53 (2014) 317–324

Contents lists available at SciVerse ScienceDirect

Materials and Design

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

Evaluation of permanent deformation characteristics of unmodifiedand Polyethylene Terephthalate modified asphalt mixtures usingdynamic creep test

0261-3069/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.matdes.2013.07.015

⇑ Corresponding author. Tel.: +60 108927064; fax: +60 379552182.E-mail addresses: p.baghaee@gmail.com, payam_baghaei@siswa.um.edu.my (T.

Baghaee Moghaddam).

Taher Baghaee Moghaddam ⇑, Mehrtash Soltani, Mohamed Rehan KarimCenter for Transportation Research, Department of Civil Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia

a r t i c l e i n f o

Article history:Received 18 March 2013Accepted 6 July 2013Available online 15 July 2013

Keywords:Waste Polyethylene TerephthalateModified asphalt mixturePermanent deformationStress levelTemperature

a b s t r a c t

One of the major types of plastics that can be found in Municipal Solid Waste (MSW) is PolyethyleneTerephthalate (PET) which is a non-biodegradable semi-crystalline thermoplastic polymer, and is consid-ered as polyester material. Generating large amount of waste PET, mainly as bottles, would cause envi-ronmental hazards by disposing in landfills. This paper aims to evaluate effects of utilizing waste PETflakes as modifier in asphalt mixture as an alternative solution to overcome the potential risks arise fromproducing large amount of waste PET as well as evaluating the deformation characteristics of unmodifiedand PET modified asphalt mixtures. To achieve this aim, different percentages of PET were designated forthis investigation, namely: 0%, 0.2%, 0.4%, 0.6%, 0.8% and 1% by weight of aggregate particles, and dynamiccreep test was performed at different stress levels (300 kPa and 400 kPa) and temperatures (10 �C, 25 �Cand 40 �C). Consequently, Zhou three-stage model was developed. The results showed that permanentdeformation characteristics of asphalt mixture were considerably improved by utilization of PET modifi-cation, when the permanent strain was remarkably decreased in PET modified mixture compared to theconventional mixture at all stress levels and temperatures. Besides, based on Zhou model, it was con-cluded that elastic and visco-elastic properties of asphalt mixture were improved by application of PETmodification.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction temperature due to asphalt’s visco-elastic properties. Asphalt vis-

Plastics have prominent utilization worldwide. PolyethyleneTerephthalate (PET) is one of the major types of plastics that canbe found in the Municipal Solid Waste (MSW) [1]. MSW containshigh proportion of thermoplastic polymers which has beenincreasing in value. PET is a semi-crystalline thermoplastic poly-mer, and is considered as polyester material [2]. Large amount ofwaste PET is being produced in the form of different products,for instance, bottles, fibers, molding and sheets are the productsmostly manufactured in Europe by application of PET [3]. Utiliza-tion of PET is well-known in food industry because of superlativecharacteristics offered by PET as a packaging material, mainly asbottles [4].

During the last decades, number of vehicles on roads, especiallyheavy vehicle such as vans and trucks, has been increased, consid-erably. By rising the number and weight of vehicles, amount andcycles of applied loads to the asphalt mixture has increased signif-icantly, thus service life of asphalt mixture was decreased. More-over, performance of asphalt mixture is influenced by ambient

cosity is highly affected by temperature when higher temperatureresults in softer and less viscose asphalt binder which can affectthe permanent deformation performance of asphaltic concrete innegative manner.

To overcome these problems and make a better mixture provid-ing longer service life, modifying asphalt mixture by utilizing addi-tives such as fibers and polymers comes into popularity amongroad engineers and designers [5]. Among these additives, polymers(such as SBS, LDPE, and HDPE) have a prominent utilization [6–8].In this case, many investigations have been conducted on effective-ness of using waste materials as additives in order to: (1) preventfrom imposed additional charges by using additives and, (2) find asolution to reuse waste materials as secondary materials in suchconstruction projects [9–12].

Hence, this study aims to assess effects of adding waste PET asan additive on permanent deformation characteristics of asphaltmixture.

2. Background

Progressive accumulation of permanent deformation is definedas rutting. Rutting, which is caused by repeated traffic loading, issum of the total deformation accrues in each layer of pavement

318 T. Baghaee Moghaddam et al. / Materials and Design 53 (2014) 317–324

structure and, moreover, asphalt layer has shown a prominentmagnitude in rutting [13]. The permanent deformation of asphaltmixture is greatly influenced by type of asphalt mixture and per-centage of voids in mix [5]. In this case, it was reported that stonemastic asphalt (SMA) mixture has better performance in compari-son with conventional dense graded mixture [14,15].

Different test methods can be used to evaluate permanentdeformation of asphalt mixture such as static and dynamic creeptests as well as wheel tracking test among which static test wasestimated that cannot be a true indicator of permanent deforma-tion of modified mixture; however, dynamic creep test is consid-ered as one of the best test procedure to assess the permanentdeformation of asphalt mixture [16,17]. Concept of this methodwhich is based on axial compression was developed in 1970 byMonismith et al.[18]. Further, it is reported that among all test pro-cedures dynamic creep test is an appropriate laboratory method toinvestigate permanent deformation performance of modified andunmodified asphalt layers [19].

Several permanent deformation models were proposed since1970s that include power-law model, VESYS model, Ohio Statemodel, superpave and AASHTO 2002 models [18,20,21]. In past re-searches, Flow Number (FN) was a popular indicator to assess thepermanent deformation of hot mix asphalt [22,23]. This value canbe obtained as result of dynamic test by plotting cumulative plasticstrain vs number of load cycles.

There are three phases of flow during the test namely: primary,secondary and tertiary stages. During the primary stage rate ofstrain decreases by load application. In secondary phase, rate ofstrain almost remains constant, and during tertiary phase thestrain rate rises dramatically which is due to significant deforma-tion of specimen under cyclic load application. In this criterion,the FN is defined as the load cycles when the tertiary stage is ini-tiated. Although this method is one of the criteria to assess ruttingperformance of asphalt mixture, it has a disadvantage because it isobserved that several minimum values from the strain rate werefound [20].

In later investigation, in 2004, another model was developed byZhou et al. [22]. Zhou proposed three-stage model with one modelfor each stage of creep curve namely: power-law model, linearmodel and exponential model each for primary, secondary and ter-tiary stages, respectively (Fig. 1). The models are presented atfollows:

First stage : ep ¼ aNb;N 6 NPS ð1Þ

Second stage : ep ¼ ePS þ cðN � NPSÞ; and ePS ¼ aNbPS; NPS

6 N 6 NST ð2Þ

Fig. 1. Zhou three-stage model.

Third stage : ep ¼ eST þ dðef ðN�NST Þ�1Þ; and eST

¼ ePS þ cðNST � NPSÞ;NST 6 N ð3Þ

where NPS is the number of load cycle corresponding to the initia-tion of secondary stage. NST is number of load cycle correspondingto the initiation of tertiary stage. This value is also refers to FN.ePS is cumulative permanent strain at initiation of secondary stage.eST is cumulative permanent strain at initiation of tertiary stage, anda, b, c, d and f are material constants.

Zhou model was seemed to have a better correlation with per-manent deformation of asphalt mixture in field compared to othermodels. Further, this model outweighs the other models becauseeach transition points in creep curve can be obtained by using thiscriterion. Thus, in other investigation Khodaii and Mehrara [13]used this model to evaluate permanent deformation of SBS modi-fied asphalt mixture using dynamic creep test.

3. Objectives and experimental procedure

This study attempts to evaluate effects of using non-biodegrad-able PET as modifier on permanent deformation characteristics ofStone Mastic Asphalt (SMA) mixture. To achieve this aim differentpercentages of PET were designated for this investigation and dy-namic creep test was used at two different stress levels (300 and400 kPa). Moreover, to have a better understanding about PETmodification, dynamic creep test was conducted at three differenttemperatures namely: 10, 25, and 40 �C. Consequently, Zhou three-stage model was designated for each stress levels andtemperatures.

3.1. Materials

Granite-rich aggregate used in this study was prepared fromKajang Rock Quarry in Malaysia. Aggregate particle size distribu-tion of selected gradation is presented in Fig. 2.

SMA mixtures have been prepared with fresh 80/100 penetra-tion grade asphalt cement. No stabilizing additives were added toasphalt cement. In order to minimize complications, the asphaltand aggregate sources were kept the same during the study. Table 1illustrates the properties of materials utilized in this researchprogram.

3.2. Polyethylene Terephthalate (PET)

In this study, waste PET flakes were obtained from PET bottles.After collection of PET bottles, they were washed and cut to small

0

20

40

60

80

100

120

0.06 0.13 0.25 0.50 1.00 2.00 4.00 8.00 16.00

Per

cent

age

Pas

sing

(%

)

Aggregate particle size distribution

Max Limit

Min limit

Used gradation

Fig. 2. Aggregate distribution on gradation chart.

Table 1Material properties.

Property Unit Used standard Value Standardrequirement

Coarse aggregateL.A. Abrasion % ASTM: C131 19.45 <30Flakiness index % BS EN 933-3

[24]2.72 <20

Elongation index % BS 812-105.2[25]

11.26 <20

Aggregate crushingvalue

% BS 812-110[26]

19.10 <30

Bulk specific gravity – ASTM: C127 2.60 –Absorption % ASTM: C127 0.72 <2

Fine aggregateBulk specific gravity – ASTM: C128 2.63 –Absorption % ASTM: C128 0.4 <2Soundness loss % ASTM: C88 4.1 <15

AsphaltPenetration at 25 �C 0.1 mm ASTM: D5 87.5 –Softening point �C ASTM: D36 46.6 –Flash point �C ASTM: D92 300 –Fire point �C ASTM: D92 320 –Specific gravity – ASTM: D70 1.03 –

Table 2Particles size distribution of waste PET.

Sieve size (mm) Passing (%)

2.36 1001.18 270.425 5

Table 3Properties of PET.

Property Unit Method Value

Water absorption % ASTM: D570 0.11Tensile strength Psi ASTM: D638 11,500Approx glass transition temperature �C – 75Approx melting temperature �C – 250Density g/cm3 ASTM: D792 1.35

Table 4Summary of mix design.

PET (%) BSGa VMAb (%) VFAc (%) OACd (%)

0 2.294 18.12 77.92 6.770.2 2.297 17.53 77.13 6.450.4 2.297 17.47 77.10 6.430.6 2.296 17.20 76.69 6.290.8 2.291 17.30 76.89 6.361 2.283 17.55 77.20 6.51

a Bulk specific gravity of compacted mixture.b Void in mineral aggregate.c Void filled with asphalt.d Optimum asphalt content.

Fig. 3. Universal Testing Machine (UTM).

T. Baghaee Moghaddam et al. / Materials and Design 53 (2014) 317–324 319

parts. These small parts were further crushed. The crushed PETparticles were sieved, and those passed sieve 2.36 mm were cho-sen for this investigation. Different percentages of crushed PETwere used in this study, namely: 0%, 0.2%, 0.4%, 0.6%, 0.8%, and1% by weight of aggregate particles. The particle size distributionand properties of waste PET are shown in Tables 2 and 3,respectively.

3.3. Sample preparation

Three identical samples were prepared for each percentage ofPET and for each stress levels and temperatures. For preparingthe entire samples 1100 g of mixed aggregate particles and fillerwas heated at the temperature of 160 �C for 3 h. Besides, asphaltcement was heated at the temperature of 130 �C before mixingwith aggregate particles in order to achieve proper viscosity. PETparticles were added directly to mixture as the method of dry pro-cess, and 50 blows of compaction effort was considered on eachside of specimen. All conventional and modified samples were pre-pared at Optimum Asphalt Content (OAC) according to ASTM:D1559 [8,13,17], and the results are presented in Table 4.

3.4. Dynamic creep test

When a load is applied to the surface of asphalt pavement, it de-forms although a majority of deformation recovers after the load isremoved. This phenomenon was referred to the characteristics ofasphalt which known as visco-elastic material. Nevertheless, aminute amount of irrecoverable viscous deformation would stillremains in mixture. By applying millions of load cycles, the smallamounts of deformation would be accumulated and eventually re-sult in surface rutting. The amount of this deformation is greatlyinfluenced by amount and number of cyclic loads as well as envi-ronmental temperature.

Hence, the uniaxial repeated-load creep test was designated tocharacterize rutting performance of control and PET modified as-phalt mixtures. Universal Testing Machine (UTM) was utilizedwhich is the most commonly used device to measure the perma-nent deformation of asphalt mixture (see Fig. 3) [7]. During the testcyclic load with repeating time of 1000 ms (pulse compressive loadof 100 ms with rest period of 900 ms) was applied. Prior to conductthe test, all specimens were placed at controlled temperaturechamber for 3 h to reach the uniform temperature.

During the test the vertical deformation of specimen was mea-sured by Linear Variable Differential Transducers (LVDTs) which

0

50

100

150

200

250

300

350

400

450

0 5000 10000 15000 20000 25000

Cum

ulat

ive

Per

man

ent

Str

ain

(µs)

Load Cycle

Control

0.2% PET

0.4% PET

0.6% PET

0.8% PET

1% PET

Fig. 4. Cumulative permanent strain vs load cycle for control and PET modified mixtures at 300 kPa stress and 10 �C.

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0 5000 10000 15000 20000 25000

Cum

ulat

ive

Per

man

ent

Stra

in (

µs)

Load Cycle

Control

0.2% PET

0.4% PET

0.6% PET

0.8% PET

1% PET

Fig. 5. Cumulative permanent strain vs load cycle for control and PET modified mixtures at 300 kPa stress and 25 �C.

0

2000

4000

6000

8000

10000

12000

0 1000 2000 3000 4000 5000 6000 7000 8000

Cum

ulat

ive

Per

man

ent

Stra

in (

µs)

Load Cycle

Control

0.2% PET

0.4% PET

0.6% PET

0.8% PET

1% PET

Fig. 6. Cumulative permanent strain vs load cycle for control and PET modified mixtures at 300 kPa stress and 40 �C.

320 T. Baghaee Moghaddam et al. / Materials and Design 53 (2014) 317–324

were installed in vertical direction, and strain can be calculated byusing the following equation:

e ¼ hH0

ð4Þ

where e is the accumulated strain, h is the axial deformation, mm;and H0 is the initial specimen height, mm.

The test would be terminated until accumulative load cyclesreach 20,000, or until a large deformation caused the LVDTs to goout of range [22].

4. Results and discussion

This section discusses about permanent deformation perfor-mance of control (unmodified) and PET modified asphalt mixtures.

0

200

400

600

800

1000

1200

1400

1600

0 5000 10000 15000 20000 25000

Cum

ulat

ive

Per

man

ent

Stra

in (µ

s)

Load Cycle

Control

0.2% PET

0.4% PET

0.6% PET

0.8% PET

1% PET

Fig. 7. Cumulative permanent strain vs load cycle for control and PET modified mixtures at 400 kPa stress and 10 �C.

0

1000

2000

3000

4000

5000

6000

7000

0 5000 10000 15000 20000 25000

Cum

ulat

ive

Per

man

ent

Stra

in (

s)

Load Cycle

Control

0.2% PET

0.4% PET

0.6% PET

0.8% PET

1% PET

Fig. 8. Cumulative permanent strain vs load cycle for control and PET modified mixtures at 400 kPa stress and 25 �C.

0

2000

4000

6000

8000

10000

12000

0 500 1000 1500 2000 2500 3000 3500

Cum

ulat

ive

Per

man

ent

Stra

in (

s)

Load Cycle

Control

0.2% PET

0.4% PET

0.6% PET

0.8% PET

1% PET

Fig. 9. Cumulative permanent strain vs load cycle for control and PET modified mixtures at 400 kPa stress and 40 �C.

T. Baghaee Moghaddam et al. / Materials and Design 53 (2014) 317–324 321

Three types of strains, namely: perfectly recoverable elastic, visco-elastic and irrecoverable plastic strains may develop after applyinga load to the asphalt mixture. At higher temperature or underslow-moving loads, asphalt mixture may have pure elastic or vis-coelastic behavior; however, it might behave as an elastic brittlematerial at lower temperature or under fast-moving loads [27].

4.1. Permanent strain

Total permanent strain of control and PET modified SMA mix-tures were calculated and compared with each other. The resultsare illustrated in Figs. 4–9. As can be seen in these figures, ruttingresistance of PET modified mixture improved remarkably com-

Table 5Three stage models and boundary points for control and PET modified mixtures at 300 kPa stress.

Temperature (�C) PET (%) Primary stage Secondary stage Tertiary stage

Model End point Model End point (FN) Model

100 ep = 9.2861N0.3758 a a a a

0.2 ep = 8.5225N0.3527 a a a a

0.4 ep = 20.7908N0.32534 a a a a

0.6 ep = 5.2546N0.3530 a a a a

0.8 ep = 2.0610N0.4350 a a a a

1 ep = 4.7316N0.2891 a a a a

250 ep = 237.6292N0.2063 a a a a

0.2 ep = 277.8297N0.1836 a a a a

0.4 ep = 168.5885N0.2110 a a a a

0.6 ep = 95.0441N0.2516 a a a a

0.8 ep = 62.1749N0.2696 a a a a

1 ep = 7.6634N0.4731 a a a a

400 ep = 25.3780N0.5998 1117 ep = 1708.76 + 1.6277(N � 1117) 2637 ep ¼ 4185:12þ 9540:21ðe0:0003ðN�2637Þ � 1Þ0.2 ep = 71.8700N0.4355 1513 ep = 1743.28 + 1.1169(N � 1513) 3033 ep ¼ 3377:10þ 5610:02ðe0:0004ðN�3033Þ � 1Þ0.4 ep = 66.4529N0.4654 2001 ep = 2237.33 + 1.2336(N � 2001) 3833 ep ¼ 4540:43þ 6409:98ðe0:0004ðN�3833Þ � 1Þ0.6 ep = 141.9040N0.3609 2020 ep = 2286.44 + 0.9045(N � 2020) 4233 ep ¼ 4261:21þ 6940:13ðe0:0004ðN�4233Þ � 1Þ0.8 ep = 54.0753N0.4802 2393 ep = 2323.10 + 1.0099(N � 2393) 4633 ep ¼ 4486:39þ 7939:89ðe0:0003ðN�4633Þ � 1Þ1 ep = 59.9830N0.4585 2473 ep = 2211.34 + 0.9221(N � 2473) 4873 ep ¼ 4345:25þ 8240:11ðe0:0002ðN�4873Þ � 1Þ

a Not found at the end of 20,000 load cycle.

Table 6Three stage models and boundary points for control and PET modified mixtures at 400 kPa stress.

Temperature (�C) PET (%) Primary stage Secondary stage Tertiary stage

Model End point Model End point (FN) Model

100 ep = 210.9134N0.1872 a a a a

0.2 ep = 72.2730N0.2775 a a a a

0.4 ep = 19.5347N0.3478 a a a a

0.6 ep = 10.6882N0.3848 a a a a

0.8 ep = 26.5544N0.2544 a a a a

1 ep = 13.4996N0.2137 a a a a

250 ep = 19.3424N0.5452 4625 ep = 1969.66 + 0.4276(N � 4625) 10,065 ep = 4247.35 + 7320.02(e0.0001(N – 10,065) � 1)0.2 ep = 30.5465N0.4400 8609 ep = 1682.16 + 0.1723(N � 8609) a a

0.4 ep = 31.3203N0.4324 11,169 ep = 1809.41 + 0.1012(N – 11,169) a a

0.6 ep = 63.4469N0.3522 16,001 ep = 1980.63 + 0.0718(N – 16,001) a a

0.8 ep = 8.7548N0.5426 a a a a

1 ep = 9.3355N0.5167 a a a a

400 ep = 42.7569N0.7009 359 ep = 2677.66 + 8.3813(N � 359) 819 ep = 6612.33 + 4980.11(e0.0024(N�819) � 1)0.2 ep = 28.6207N0.6916 519 ep = 2201.72 + 4.5453(N � 519) 1219 ep = 5584.42 + 8070.01(e0.0010(N�1219) � 1)0.4 ep = 19.4108N0.7261 659 ep = 2205.24 + 4.4332(N � 659) 1339 ep = 5326.27 + 8649.96((e0.0009(N�1339) � 1)0.6 ep = 27.8015N0.6675 797 ep = 2439.57 + 3.4728(N � 797) 1557 ep = 4954.23 + 5150.01((e0.0011(N�1557) � 1)0.8 ep = 42.6260N0.5911 877 ep = 2384.68 + 3.0153(N � 877) 1717 ep = 4789.72 + 5750.17((e0.0008(N�1717) � 1)1 ep = 10.2958N0.7537 1237 ep = 2251.92 + 2.7001(N � 1237) 2277 ep = 4886.42 + 7169.80((e0.0007(N�2277) � 1)

a Not found at the end of 20,000 load cycle.

322 T. Baghaee Moghaddam et al. / Materials and Design 53 (2014) 317–324

pared to the control mixture, and, in all cases, total permanentstrain decreases at higher PET contents and for a certain numberof load application while adding 1% PET results in the lowestamount of strain.

In previous research [9] which was conducted on evaluation ofdifferent polymer modified asphalt, because of high melting pointof PET it was supposed that PET hinders the mixing, and was notsuitable to be used in asphalt modification. Nevertheless, the resultof this investigation shows that using PET as modifier can consid-erably improve permanent deformation characteristics of asphaltmixture. Improvement of permanent deformation characteristicsof PET modified asphalt mixture can be referred to PET propertieswhich is known as semi-crystalline material. That is to say, with

regards to the mixing temperature which is around 160 �C, and isover the glass transition temperature of PET (75 �C) the amorphousregion of PET would be changed to liquid form; however, a crystal-line region would still exist as solid part due to the high meltingpoint of PET which is around 250 �C [28,29]. Hence, it is deemedthat PET might improve asphalt mixture characteristics in twoways: first the molten and liquid part of PET strengthens the bondsbetween aggregate particles and asphalt binder, and second the so-lid region of PET absorbs or distracts some amount of energy ap-plied by cyclic load.

In previous study [30], the wheel tracking test was used to eval-uate the effect of PET on rutting resistance of asphalt mixture, andit was shown that adding 4% of PET (by weight asphalt content) re-

0

500

1000

1500

2000

2500

3000

0 0.2 0.4 0.6 0.8 1

Cyc

les

PET (%)

(A) Number of Load Cycles at First Stage

Number of Load Cycles at Second Stage

0

200

400

600

800

1000

1200

1400

0 0.2 0.4 0.6 0.8 1

Cyc

les

PET (%)

(B) Number of Load Cycles at First Stage

Number of Load Cycles at Second Stage

Fig. 10. Number of load cycles at first and second stages for control and PETmodified mixtures at 300 kPa and 400 kPa stresses and temperature of 40 �C (A:300 kPa stress, and B: 400 kPa stress).

26373033

38334233

46334873

8191219 1339

15571717

2277

0

1000

2000

3000

4000

5000

6000

0 0.2 0.4 0.6 0.8 1

FN

(C

ycle

s)

PET (%)

300 kPa 400 kPa

Fig. 11. Flow Number (FN) vs PET content at 40 �C. (FN is the point or cycle numberat which pure plastic shear deformation occurs).

T. Baghaee Moghaddam et al. / Materials and Design 53 (2014) 317–324 323

sulted in the best rutting behavior when rutting resistance of mix-ture was improved by 29% compared to the control mixture, butthe influence of using PET seems to be much lower in comparisonwith the results obtained in this study. The reason might be due tothe size of PET. In other words, in that study the maximum size ofPET particles was 1.18 mm which is half of the maximum size usedfor this investigation (2.36 mm) and because the PET particles donot melt during mixture fabrication, it can influence the mixturemechanical properties. In addition, a study on fatigue characteris-tics of PET reinforced asphalt mixture showed that using 1% PET asreinforcement can improve the fatigue life of PET modified mixtureunder dynamic loading by 124.8% [28].

It is good to note that the decrease in permanent strain is moreconsistent with previous investigation on using styrene–butadi-ene–styrene (SBS) as modifier in asphalt mixture [13]. In that studyit was concluded that the mixtures fabricated by SBS modified bin-

der had remarkably lower permanent strain and the mixture con-tained 5% of SBS showed the best mechanical behavior when after10,000 loading cycles the permanent strain decreased from10,000 ls for 5%-SBS modified mixture to under 3000 ls forunmodified mixture at temperature of 40 �C and 200 kPa stress.Further, other investigation on mineral and cellulose fibers showedthat although using these additives could improve the rutting per-formance of asphalt mixture, the effect did not seem to be signifi-cant in comparison with the SBS modification [7].

4.2. Temperature and stress level

In order to have better understanding about rutting characteris-tics of SMA mixture at different environmental temperature, dy-namic creep test was conducted at temperatures of 10, 25 and40 �C which are considered as relatively low, medium and hightemperatures, respectively. Besides stress levels of 300 kPa and400 kPa were designated for this study. In previous research, itwas reported that lower stress level (e.g. 100 kPa) cannot properlymanifest the rutting performance of modified asphalt mixtures[13].

Fig. 4 shows that the final amount of accumulated strain at10 �C and 300 kPa stress for control mixture is very low whenthe amount of strain for control mixture is under 400 ls after20,000 load application. Nevertheless, this value is much lowerfor the mixture reinforced by 1% PET which is below 100 ls. Totalpermanent strain increases at higher temperatures. For instance, atthe same stress level, in case of control mixture when the temper-ature increases from 10 �C to 25 �C (Fig. 5) the strain value rises tonearly 2000 ls which is around 5 times higher in comparison with10 �C. Moreover, at the same temperature, the strain value for themixture modified by 1% PET reaches to nearly 800 ls after 20,000load repetition which is nearly 8 times higher than the amount cal-culated for 10 �C. More importantly, as can be seen in Fig. 6 at 40 �Cand after applying a number of cyclic load all amounts of strainsreach to the 10,000 ls in which the LVDTs go out of range.

Further, by rising temperature the same trend was observed at400 kPa stress level. The strain values increase by an increment instress (from 300 kPa to 400 kPa) for both 10 �C and 25 �C temper-atures (Figs. 7 and 8). It is good to note that in 400 kPa stressand temperature of 25 �C the control specimen fails after nearly12,000 load cycle when the strain value reaches to 6000 ls. Be-sides, it can be illustrated from Fig. 9 at temperature of 40 �C num-ber of cycles that LVDTs go out of range decline (around one third)compared to the same temperature at 300 kPa stress level.

4.3. Zhou’s three-stage model

In this study, to have better understanding about permanentdeformation behavior of control and PET modified mixtures; Zhouthree-stage model was proposed at different stress levels and tem-peratures. MATLAB software was utilized for modelling each stagein order for finding the parameters as well as transition points be-tween each stage. Besides, the FN was also calculated.

The results are presented in Tables 5 and 6. Achieved resultsshow at 300 kPa stress and lower temperatures (10 �C and 25 �C),the control and modified mixtures do not go through secondarystage; however, at 300 kPa stress and temperature of 40 �C allstages can be recognized. Hence, importance of temperature canapparently be recognized on permanent deformation behavior ofasphalt mixture. In other words, road pavement is much more sus-ceptible against rutting damage in hot climate condition comparedto cold regions. More importantly, as can be seen in Table 6, at400 kPa stress and 25 �C the control mixture passes through thirdstage after 10,065 loading cycles whereas the mixtures modified by0.2%, 0.4% and 0.6% of PET just enter the second stage. Furthermore,

324 T. Baghaee Moghaddam et al. / Materials and Design 53 (2014) 317–324

the mixtures modified by 0.8% and 1% PET still remain at the firststage.

Fig. 10 manifests number of load cycles at first and secondstages at 40 �C for both 300 kPa and 400 kPa stress levels. From thisfigure it can be illustrated that lengths of first and second stages(elastic and viscoelastic stages) are increased in PET modified mix-tures when the mixtures including higher PET content showed tohave higher amount in length. At 300 kPa, for instance, length offirst stage rises from 1117 to 2473 cycles for control and 1%-PETmodified mixture which is almost two times longer, and lengthof second stage increases from 1520 to 2400. It should be notedthat effect of PET modification can be more highlighted at higherstress level, while at 400 kPa stress for mixture modified by 1%PET first and second stages are 3.45 and 2.27 times longer respec-tively compared to control mixture.

In addition, as can be seen in Fig. 11, FN increased considerablyby adding waste PET in asphalt mixture. For instance, for controland 1%-PET modified mixtures the FN rise from 2637 to 4873and 819 to 2277 cycles for 300 kPa and 400 kPa stress, respectively.

5. Conclusions

In this investigation, permanent deformation characteristics ofcontrol and PET modified asphalt mixture were evaluated usingdynamic creep test. Dynamic creep test was conducted at differenttemperatures and stress levels. Creep models were proposed, andthe results are summarized as follows:

(1) Waste PET showed to have great potential to be reused asmodifier in asphalt mixture.

(2) Permanent deformation characteristics of PET modifiedasphalt mixtures remarkably improved compared to thecontrol mixture, and the mixture modified by higher amountof PET showed to have better resistance against permanentdeformation.

(3) At higher stress level and temperature, permanent deforma-tion resistance of both control and PET modified mixturedecreased.

(4) Based on Zhou three-stage model, length of first (elastic) andsecond (viscoelastic) stage of creep curve increased consid-erably in PET modified mixture, and effect of PET modifica-tion was more highlighted at higher stress level (400 kPa).

(5) At temperature of 40 �C and at both 300 kPa and 400 kPastress level, Flow Number (FN) was considerably increasedby application of PET modification.

Acknowledgement

This investigation was supported by scientific research fundprovided by Malaysian Ministry of Higher Education (Project num-ber: FP021-2011A).

References

[1] Li Y, White DJ, Lee Peyton R. Composite material from fly ash and post-consumer PET. Resour Conserv Recycl 1998;24:87–93.

[2] Shukla SR, Harad AM. Aminolysis of polyethylene terephthalate waste. PolymDegrad Stabil 2006;91:1850–4.

[3] Torres N, Robin JJ, Boutevin B. Study of thermal and mechanical properties ofvirgin and recycled poly(ethylene therephtalate) before and after injectionmolding. Eur Polym J 2000;36:2075–80.

[4] Navarro R, Ferrándiz S, López J, Seguí VJ. The influence of polyethylene in themechanical recycling of polyethylene terephtalate. J Mater Process Technol2008;195:110–6.

[5] Moghaddam TB, Karim MR, Abdelaziz M. A review on fatigue and ruttingperformance of asphalt mixes. Sci Res Essays 2011;6:670–82.

[6] Tayfur S, Ozen H, Aksoy A. Investigation of rutting performance of asphaltmixtures containing polymer modifiers. Constr Build Mater 2007;21:328–37.

[7] Mokhtari A, Moghadas Nejad F. Mechanistic approach for fiber and polymermodified SMA mixtures. Constr Build Mater 2012;36:381–90.

[8] Al-Hadidy AI, Yi-qiu T. Mechanistic approach for polypropylene-modifiedflexible pavements. J Mater Des 2009;30:1133–40.

[9] Casey D, McNally C, Gibney A, Gilchrist MD. Development of a recycledpolymer modified binder for use in stone mastic asphalt. Resour ConservRecycl 2008;52:1167–74.

[10] Hınıslıoglu S, Agar E. Use of waste high density polyethylene as bitumenmodifier in asphalt concrete mix. Mater Lett 2004;58:267–71.

[11] Arabani M, Mirabdolazimi SM, Sasani AR. The effect of waste tire thread meshon the dynamic behaviour of asphalt mixtures. Constr Build Mater2010;24:1060–8.

[12] Asaad JN, Tawfik SY. Polymeric composites based on polystyrene and cementdust wastes. J Mater Des 2011;32:5113–9.

[13] Khodaii A, Mehrara A. Evaluation of permanent deformation of unmodifiedand SBS modified asphalt mixtures using dynamic creep test. Constr BuildMater 2009;23:2586–92.

[14] Kennepohl GJ, Davidson JK. Introduction of stone mastic asphalt in Ontario. JAssoc Asphalt Paving Technol 1992;61:517–34.

[15] Brown ER, Mallick RB, Haddock JE, Bukowski J. Performance of stone matrixasphalt (SMA) mixtures in the United States. National Center for AsphaltTechnology Report No. 97-1. Alabama: Auburn University; 1997.

[16] Valkering CP, Lancon DJ, de Hilster E, Stoker DA. Rutting resistance of asphaltmixes containing non-conventional and polymer modified binders. J AssocAsphalt Paving Technol 1990;59:590–609.

[17] Kalyoncuoglu SF, Tigdemir M. A model for dynamic creep evaluation of SBSmodified HMA mixtures. Constr Build Mater 2011;25:859–66.

[18] Monismith CL, Ogawa N, Freeme CR. Permanent deformation of subgrade soilsdue to repeated loadings. Transport Res Rec 1975;537:1–17.

[19] Kaloush KE, Witczak MW. Tertiary flow characteristics of asphalt mixtures. JAssoc Asphalt Paving Technol 2002;71:248–80.

[20] Goh SW, You Z. A simple stepwise method to determine and evaluate theinitiation of tertiary flow for asphalt mixtures under dynamic creep test.Constr Build Mater 2009;23:3398–405.

[21] Zhou F, Scullion T. Preliminary field validation of simple performance testsfor permanent deformation: case study. Transport Res Rec 2003;1832:209–16.

[22] Zhou F, Scullion T, Sun L. Verification and modeling of three-stage permanentdeformation behavior of asphalt mixes. J Transp Eng-ASCE 2004;130:486–94.

[23] Witczak MW. Specification criteria for simple performance tests for rutting.Volume I: Dynamic modulus (E⁄) and Volume II: Flow number and flow time.National Cooperative Highway Research Program Report No: 580,Transportation Research Board, Washington DC; 2007.

[24] BS EN 933-3:2012. Tests for geometrical properties of aggregates.Determination of particle shape. Flakiness index. London, UK: BritishStandards Institution; 2012.

[25] BS 812-105.2:1990. Methods for determination of particle shape.Elongation index of coarse aggregate. London, UK: British StandardsInstitution; 1990.

[26] BS 812-110:1990. Testing aggregates. Methods for determination of aggregatecrushing value (ACV). London, UK: British Standards Institution; 1990.

[27] Al-Qadi IL, Yoo PJ, Elseifi MA, Nelson S. Creep behavior of hot-mix asphalt dueto heavy vehicular tire loading. J Eng Mech 2009;135:1265–73.

[28] Baghaee Moghaddam T, Karim MR, Syammaun T. Dynamic properties of stonemastic asphalt mixtures containing waste plastic bottles. Constr Build Mater2012;34:236–42.

[29] Ahmadinia E, Zargar M, Karim MR, Abdelaziz M, Shafigh P. Using waste plasticbottles as additive for stone mastic asphalt. J Mater Des 2011;32:4844–9.

[30] Ahmadinia E, Zargar M, Karim MR, Abdelaziz M, Ahmadinia E. Performanceevaluation of utilization of waste Polyethylene Terephthalate (PET) in stonemastic asphalt. Constr Build Mater 2012;36:984–9.