Evaluation of rutting, fatigue and moisture damage performance of nanoclay modified asphalt binder

Construction and Building Materials 113 (2016) 341–350

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Evaluation of rutting, fatigue and moisture damage performance of nanoclay modified asphalt binder Prabin Kumar Ashish a, Dharamveer Singh a,⇑, Siva Bohm b a b

Department of Civil Engineering, Indian Institute of Technology Bombay, Mumbai 400076, India Department of Metallurgical Engineering and Materials Science, Indian Institute of Technology, Mumbai 400076, India

h i g h l i g h t s
Nanoclay (CL-30B) addition to binder showed improved rutting resistivity potential.
SFE approach showed improved moisture resistivity of binder with CL-30B addition.
LAS study showed improvement in fatigue life of binder with CL-30B modification.

a r t i c l e

i n f o

Article history: Received 26 November 2015 Received in revised form 27 February 2016 Accepted 14 March 2016

Keywords: CL-30B Rutting Moisture damage SFE Fatigue life LAS

a b s t r a c t Recent impetus on utilization of different types of nanomaterials for modification of asphalt binders has motivated the authors to undertake the present study. The present study evaluated the rutting, fatigue and moisture damage performance of nanoclay (CL-30B) modified asphalt binder based on newly adopted test methods. Based on Superpave rutting parameter, it was observed that rutting resistivity of a binder increases with an increase in CL-30B content. Moisture resistivity of CL-30B modified asphalt binder with different types of aggregates system was studied using Surface Free Energy (SFE) approach. The SFE components of nanoclay modified binder were measured using Wilhelmy plate method. Four different types of aggregates (Basalt, Limestone, Sandstone and Granite) were chosen for the study. Overall increase in total SFE of the binder was observed with addition of CL-30B. Increase in work of cohesion and decrease in work of debonding was observed with an increase in CL-30B for all type of considered aggregate. Based on Energy ratio (ER), asphalt binder with basaltic aggregate was found to have better moisture damage resistivity among different types of aggregate selected in this study. The fatigue performance of CL-30B modified binder was evaluated using Linear Amplitude Sweep (LAS) test which is based on Visco Elastic Continuum Damage (VECD) theory. The analyses of the data showed that addition of CL-30B enhances fatigue life of a binder. The study showed potential of CL-30B to enhance various rheological performance of a binder for better and long lasting pavements. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction There has been increasing interest in highway community on improving quality of asphalt binders to enhance design life of flexible pavements. Though polymer modified binders have been immensely popular, the high cost and thermal instability of these binders encouraged researchers to explore new materials to improve performance of binders. Recent impetus on utilization of different types of nanomaterials namely, nanoclay, nanosilica, nanozinc oxide, and nanolime for modification of asphalt binders ⇑ Corresponding author. E-mail addresses: [email protected] (P.K. Ashish), [email protected] (D. Singh), [email protected] (S. Bohm). http://dx.doi.org/10.1016/j.conbuildmat.2016.03.057 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.

has motivated the authors to undertake the present study [1–4]. Hossain et al. [5] reported that cost of nanoclay modified asphalt binder can be approximately 22–33% lower than that of polymer modified asphalt binder. The studies conducted by numerous researchers showed that organo modified montmorillonite nanoclay can be a potential solution to minimize rutting failure [6– 8]. The Superpave rutting parameter provides a valuable information on rutting resistant of a binder; however, limited work has been reported on nanoclay modified asphalt binders. Similarly, effect of nanoclay on fatigue performance of asphalt binders is evaluated by some researchers using different approaches. For instance, fatigue life of asphalt binder containing nanoclay was evaluated by Liu et al. [7] and Wu et al. [9] using stress controlled mode and improvement in fatigue life was

342

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

reported. Similarly, Yu et al. [10] reported improvement in fatigue life of asphalt binder with addition of nanoclay evaluated using Superpave fatigue parameter. Contrary to this, Jahromi et al. [11] and Ghile [12] observed reduction in fatigue life of binders with addition nanoclay evaluated using Superpave fatigue parameter. Ghaffarpour et al. [13] showed dependency of fatigue life of nanoclay modified asphalt mixture on temperature. In this study, improvement in fatigue life of nanoclay modified mixture was observed at 25 °C, while reduced fatigue life was reported at 5 °C compared to control mix. Hintz et al. [14] reported that determination of fatigue life of asphalt binder using under stress controlled test is time consuming and usually unrepeatable. Similarly, Superpave fatigue parameter is estimated by conducting test under linear visco-elastic range of asphalt binder and hence it lacks to capture the damage beyond this range. Also, it doesn’t account for different traffic loading experienced in the actual situation [14–15]. Recently linear amplitude sweep (LAS) test is developed based on viscoelastic continuum damage (VECD) approach for better characterization of fatigue life of asphalt binders. The LAS has been reported as a promising test to evaluate the fatigue behaviour of asphalt binders under high stress and strain conditions beyond their viscoelastic range [14–16]. So far, as per the authors’ understanding, limited studies have been found to utilize the LAS test to evaluate fatigue life of asphalt binder with addition of nanoclay [17]. Hence, the present study provides a valuable addition to the current repository of information available to the scientific community. Recently, Golestani et al. [18] reported improvement in moisture damage resistivity of polymer modified binders with addition of nanoclay using Tensile Strength Ratio (TSR) test. Though TSR test is being widely popular, outcome from this test showed a poor correlation with field performance of pavements and failed to address specific failure mechanism related to moisture damage of asphalt mixes [19]. An important fundamental material property which can help to address moisture resistivity of asphalt mixes is the surface free energy (SFE) of aggregate and asphalt binder. The SFE approach has been reported as a vital tool to evaluate the moisture damage potential of conventional and polymer modified asphalt binders by various researchers [20–23]. Recently, Hamedi et al. [3] used the SFE approach to study the moisture resistivity of nanozinc oxide modified asphalt binder. Improvement in the total SFE was observed with addition of nanozinc oxide. Similarly, Hossain et al. [5] evaluated moisture damage potential of a PG64-22 binder modified with Cloisite-15A and Cloisite-11B using the SFE approach. Increase in the total SFE was observed with addition of nanoclay. However, decrease in compatibility between aggregate and asphalt binder was observed with addition of nanoclay indicating a poorer moisture resistive damage potential of nanoclay modified asphalt binder. Overall, the review of literature showed mixed trend of moisture damage resistivity of nanoclay modified binders. Also, it has not been extensively evaluated using the SFE approach. Considering limited work reported on characterization of nanomaterials modified binders, the present study focuses on evaluation of rutting, fatigue and moisture damage potential of nanoclay modified asphalt binder using some recently developed advanced test methods and approaches. It is expected that present study would help in developing better understanding on characterization of nanoclay modified asphalt binders. 2. Objectives The specific objectives of the present research study are to: a. Evaluate rutting behaviour of nanoclay modified asphalt binders using Superpave rutting factor.

b. Evaluate effects of nanoclay on moisture susceptibility of asphalt binder and aggregate system using SFE approach. c. Assess fatigue performance of asphalt binders with and without nanoclay using LAS test.

3. Theoretical background 3.1. Review on SFE The SFE is defined as the magnitude of energy required to increase the unit surface area under vacuum condition [24]. The Acid-Base theory proposed by Van Oss et al. [24] divided total energy (kT) into three components as (a) Van der Walls/apolar (klw), (b) Lewis/monopolar acid (k+) and (c) Lewis/monopolar base (k) as expressed in Eq. (1).

pffiffiffiffiffiffiffiffiffiffiffi kT ¼ klw þ 2 kþ k

ð1Þ

Using acid-base theory, work of adhesion (DWdry) between two materials can be expressed in terms of their respective SFE components as given in Eq. (2).

qffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffi A B DW dry ¼ 2 klw klw þ kþA kB þ kA kBþ

ð2Þ

where, A and B represent aggregate and asphalt binder used respectively. A higher value of work of adhesion indicates that higher energy will be required to create unit area of new surface in dry condition. Likewise, in wet condition, water tries to displace asphalt from aggregate surface, and creates two new interfaces (i.e., water aggregate and asphalt-water), resulting in release of energy. Based upon the concept of interfacial energy, external work required for displacing binder from aggregate-binder interface is ‘‘kAB”. Similarly, work done for formation of two new surfaces is ‘‘kWA + kWB”. The total work done to displace the asphalt binder from aggregate surface is known as work of debonding (DWwet) as shown in Eq. (3). A lesser value of DWwet is desirable to have moisture resistant mix [22].

DW wet ¼ kWA þ kWB  kAB qffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi W B kW ¼ 2kW kW kBlw kW kBþ kW lw þ 4 þ k  2 lw  2 þ k  2  qffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi A W A A  2 klw klw  2 kW kþA kW kBlw klw þ2 þ k  2  þ2 qffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffi  kBþ kA þ 2 kA kþA

ð3Þ

where, Superscript W, A and B represent water, aggregate and asphalt binder respectively. Considering work of adhesion and work of debonding, Little et al. [19] defined a single term known as ‘‘Energy Ratio (ER)” which quantifies moisture damage potential of asphalt mixes as shown in Eq. (4)

  DW  W BB    dry ER ¼     DW wet

ð4Þ

where, WBB represents work of cohesion of asphalt binder. Asphaltic mixture with higher ER value will have better moisture resistant potential and vice versa. Based on ER value, Bhasin et al. [21] recommended the following moisture resistant categories of asphalt mixes (a) good:[A], when ER > 1.5; (b) fair:[B], when 0.75 < ER < 1.5; (c) poor:[C], when 0.5 < ER < 0.75; and (d) very poor:[D], when ER < 0.5.

343

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

3.2. Fatigue characterization using visco-elastic continuum damage (VECD) analysis

given in Eq. (7). This research paper uses the concept of peak shear stress in order to define the failure point.

The LAS test uses VECD theory in order to quantify the damage to asphalt binder due to repeated application of load. The major advantage of this test is that by conducting a single test, it is possible to predict the binder behaviour under varying strain level. The VECD theory has been extensively used by numerous researchers in order to study the damage analysis to binder and asphalt mixes [16,25–26]. As per AASHTO: TP 101 [27], damage to the binder at any time can be expressed as

Df ¼

DðtÞ ffi

 1=C2 C 0  C atpeakstress C1

ð7Þ

Finally, fatigue equation can be expressed as given by Eq. (8);

Nf ¼ Aðcmax ÞB

ð8Þ f ðD Þð1þð1C 2 ÞaÞ a , 2 ÞaÞðpC 1 C 2 Þ Þ

where, A ¼ ðð1þð1Cf

f = loading frequency in Hz; cmax =

maximum expected binder strain (%) and B = 2a.

N X a=1þa ½pc20 ðC i1  C i Þ ðt i  t i1 Þ1=1þa

ð5Þ

i¼1

4. Materials and experimental methodology ⁄

⁄

where, Ct = G (t)/G (Initial) = integrity parameter; c0 = applied strain level (%); G⁄ = complex modulus (MPa); a = 1/m where m is the slope of logarithmic plot between storage modulus and applied frequency; t = testing time (second). Power law model has been used in order to develop relationship between damage to the integrity parameter as given by Eq. (6) [16,27].

CðtÞ ¼ C 0  C 1 DC 2

ð6Þ

where, C1 and C2 are coefficients of curve fitting equations. Further, damage value at failure point is being calculated. Johnson [25] considered 35% reduction in integrity parameter as failure point, whereas Bahia et al. [28] has considered peak shear stress as the failure point. The damage value at failure is further calculated as Table 1 Basic properties of asphalt binder (AC-10).

4.1. Materials AC-10 grade asphalt binder was obtained from a known source for the present research study. AC-10 is a softer grade of asphalt binder which is used for preparation of different modified binder. Basic properties of AC-10 are presented in Table 1. Aggregate of different surface characteristic (Basalt, Limestone, Granite and Sandstone) having known surface energy components were selected from literature [29]. A commercially available organo modified nanoclay, CLOISITE-30B (CL-30B) was collected from Southern clay Inc.

4.2. Experimental plan Fig. 1 shows an experimental plan for the present study. The plan shows preparation of nanoclay modified binders, preparation of samples, and testing them for various performance tests. The following section provides detail information about the flow chart.

4.3. Preparation of CL-30B modified binder

Properties

Values

Standard

Softening point (°C) Penetration 0.1 mm at 25 °C Ductility at 25 °C (in cm) Absolute viscosity at 60 °C (in Poises) Flash point (°C)

45 86 79 Min. 800 Min. 220

ASTM ASTM ASTM ASTM ASTM

D36 D5 D113 D2170 D92

The CL-30B was mixed with virgin binder (AC-10) using a high shear mixer. Mixing of different doses of CL-30B (2, 4 and 6% by weight of asphalt binder) with virgin binder was carried out at 155 ± 5 °C at rotational speed of 4000 rpm for a period of 2 h as shown in Fig. 1. The control binder (0% CL-30B) was also stirred at same temperature and rotational speed for 2 h in order to maintain the similar condition as that of CL-30B modified binders. The short term aging of asphalt binder sample

Fig. 1. Experimental flow chart.

344

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350 quency sweep test and (2) amplitude sweep test. Initially, frequency sweep (0.2– 30 Hz) at constant amplitude strain within linear visco-elastic range level (0.1%) was carried out in order to get information regarding undamaged material property (a). Further, amplitude sweep test (0–30%) at frequency level of 10 Hz was conducted. During amplitude sweep test, loading amplitude was linearly increased from 0 to 30%.

was further carried out using Thin Film Oven (TFO) as per ASTM:D 1754 [30] at 163 °C for 5 h, followed by long term aging using Pressure Aged vessel (PAV) as per ASTM:D 6521[31] subjected to 2.1 MPa pressure and 100 °C for 20 h. 4.4. Dispersion characterization using X-ray diffraction (XRD) Degree of dispersion of CL-30B in binder was evaluated using the XRD technique by measuring the change in basal spacing (spacing corresponding to similar face of adjacent layer) based upon Bragg’s law (Eq. (9)). Higher value of basal spacing represents the higher interlayer spacing and visa-versa. The degree of intercalation (maintaining specific interlayer spacing) or exfoliation (interlayer spacing no longer existing) can also be examined using XRD test. Distinct peak in diffractograms represents the intercalated structure, whereas absence of distinct peak represents either non-crystalline structure or exfoliated structure. The XRD test was conducted on controlled binder as well as CL-30B modified binders by PANalyticalÒ using Cu-Ka radiation (k = 1.54 Å, Ka ratio = 0.5, Voltage = 45 kV). The XRD scanning was done in 2h range of 2–10° at a scan step size of 0.026°.

nk ¼ 2d sinðhÞ

5. Results and discussion 5.1. X-ray diffraction (XRD) analysis Fig. 2 shows the plot between 2h and intensity for controlled binder (AC-10) modified with different amount of CL-30B. Distinct peak in case of pure CL-30B, 2% and 4% CL-30B modified asphalt binder sample can be observed which shows intercalation of CL30B. Distinct peak for pure CL-30B corresponds to an interlayer spacing of 18.70 Å Also, decrease in 2h value from 3.43° to 2.99° can be observed which corresponds to increase in interlayer spacing from 25.43 Å to 29.51 Å when CL-30B increased from 2 to 4%. No distinct peak for 6% CL-30B modified asphalt binder was observed which shows the exfoliation of CL-30B structure. Further, no distinct peak in case of controlled binder (AC-10) shows the non-crystalline nature of asphalt binder.

ð9Þ

where, k = wavelength of incident X-ray, d = interlayer spacing, and h = angle of incident. 4.5. Rutting resistivity Rutting resistivity of CL-30B modified asphalt binder was examined using Superpave rutting parameter (G⁄/Sind) as per ASTM: D 7175 [32]. It gives the temperature at which binder fails in high temperature zone. As per Superpave criteria, binder will be considered as failed in rutting when G⁄/Sind value drops below 2.2 kPa for short term aged sample. High temperature grading was carried out on short term aged sample using dynamic shear rheometer (DSR) having parallel plate geometry with 25 mm diameter and 1 mm gap between them. This parameter was measured at frequency of 10 rad/s at different performance grading temperature at temperature bump of 6 °C by applying strain level under visco elastic range.

5.2. Rutting resistivity Though bulk of rutting resistance of asphalt mixes comes from the structure of the aggregate skeleton, selection of an appropriate binder also play an important role for production of stronger, stable and rut resistant mix. Variation of rutting factor on short term aged sample for different binder combinations is shown in Fig. 3. Increasing trend of rutting factor value indicates improvement in rutting resistivity with increase in CL-30B content. Significant improvement in rutting factor value can be observed when CL-30B increased from 4 to 6%. This improvement is a result of transformation of CL-30B structure from intercalated structure (4% CL-30B) to exfoliated structure (6% CL-30B). Considering 64 °C as maximum pavement temperature, increase in rutting factor value by 21%, 108% and 334% was observed when CL-30B increased from 0 to 2%, 4% and 6%, respectively. Performance grade for 0, 2, 4 and 6% CL-30B modified controlled binder can be observed to be PG70-XX, PG70-XX, PG76-XX and PG82-XX, respectively. Though, improvement in rutting factor value is observed when CL-30B increased from 0 to 2%, high performance grading remained same for both. The PG grade of control binder bumped by 1 PG and 2 PG with addition of 4% and 6% of CL-30B respectively, indicating significant improvement in high temperature grade.

4.6. SFE of asphalt binders using Wilhelmy plate method The SFE components of asphalt binders with and without CL-30 modified binders were measured using Wilhelmy Plate method as per NCHRP: 104 [19]. Samples of uniform film thickness (0.10–0.15 mm on each side of glass slide) were prepared by dipping the micro glass slide (Dimension of 24 ⁄ 50 mm, No 1.5) into molten binder samples. A uniform film thickness is very important in order to reduce the variability among replicate samples [33]. Typically, three replicate samples using three different probe liquid (Distilled water, Glycerol and Formamide) has been used for measurement of dynamic contact angle [5]. Test was conducted at controlled temperature condition of 25 °C. A speed of (0.04 mm/s) during receding and advancing phase was maintained for sample suspended with crocodile shaped clip-hook assembly. It is very important to maintain very slow speed during testing to eliminate the effect of other factors such as inertia forces and hence to maintain quasi-static condition in which inertial forces approach to zero. Also, utmost care was taken to keep the sample suspended at right angle to the base of micro balance. Total penetration depth of 10 mm was maintained in all the cases. Micro balance measures the change in weight continuously during both receding and advancing stage, and dynamic contact angles were calculated using integrated software. 4.7. LAS test The LAS test was conducted on PAV aged CL-30B modified asphalt binders using DSR on sample having 8 mm diameter and 2 mm gap between parallel plates at 25 °C as per AASHTO: TP 101[27]. This test was performed at two stages i.e.; (1) fre-

AC-10+0%CL-30B

AC-10+2%CL30B

AC-10+4%CL30B

AC-10+6%CL30B

CL-30B

Intensity

AC10+4%CL-30B

AC-10+2%CL-30B AC-10+6%CL-30B

CL-30B

AC-10+0%CL-30B

2

3

4

5

6

7

8

2θ (°) Fig. 2. XRD pattern for CL-30B modified asphalt binder.

9

10

345

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

50 45.2

AC-10+0%CL-30B AC-10+4%CL-30B

G*/Sinδ (kPa)

40

AC-10+2%CL-30B AC-10+6%CL-30B

30 21.9

20.4

20 12.7 10.4

9.8

10 4.7 0

58

9.4

5.7 2.2 2.7 64

4.6 1.1 1.3 2.2

70

4.5

76

0.6 0.7 1.1

2.2

82

Temperature (°C) Fig. 3. Variation of Superpave rutting parameter with temperature for CL-30B modified asphalt binders.

Table 2 Contact angle for CL-30B modified asphalt binder#. Sample code

AC-10 + 0%CL-30B AC-10 + 2%CL-30B AC-10 + 4%CL-30B AC-10 + 6%CL-30B

Water

Formamide

Glycerol

Average (°)

S.D

t-Statistics p = 0.05⁄

Average (°)

S.D

t-Statistics p = 0.05⁄

Average (°)

S.D

t-statistics p = 0.05⁄

95.50 96.07 99.77 98.84

0.37 0.16 0.22 0.91

Control 1.99 14.03 5.69

95.19 95.58 94.13 94.54

0.61 0.40 0.36 0.63

Control 0.757 2.12 1.148

98.04 97.50 96.66 95.02

0.59 0.55 0.80 0.32

Control 0.95 1.96 6.63

# ⁄

Number of observations for each samples with each probe liquid = 3. t-critical for p = 0.05 (2-tail) = 2.364.

5.3. SFE test 5.3.1. Variation of contact angle The contact angle can be considered as the inverse measurement of wettability. Contact angle value below 90° represents higher wettability, whereas contact angle above 90° corresponds to lower wettability [34]. Contact angle measured using three different probe liquids is presented in Table 2. Three replicates were tested for each combination. In all the cases, standard deviation value was observed to be <1°. Statistical analysis using ‘‘two tail unequal variance Student t-test” at confidence interval of 95% (p = 0.05) was carried out to check if addition of CL-30B causes significant change in the contact angle value of asphalt binders. In the case of water as a probe liquid, significant difference in contact angle can be observed for binder modified with 4% and 6% CL30B, whereas for glycerol probe liquid, significant difference in contact angle was observed only for binder modified with 6% CL30B. No significant difference in contact angle was found for formamide probe liquid. Hence, it can be concluded that degree of wettability in case of water as probe liquid is decreasing with addition of CL-30B, whereas in case of glycerol as probe liquid, wettability improved only at higher CL-30B content (6% by weight of binder). Though, decrease in contact angle is taking place with addition of CL-30B in case of formamide as probe liquid, it didn’t significantly change the wettability. Different SFE components were calculated using these contact angles and discussed in subsequent sections. 5.3.2. SFE components of asphalt binder Variation of different SFE components (acidic, basic and Liftshitz-van der Waals component) as well as total SFE value with different CL-30B content is shown in Fig. 4(a, b). Increase in acidic component and decrease in basic component of SFE can be

observed with an increase in CL-30B content. No appropriate trend was observed in the case of Van der Waal Component. Even though the basic component decreased with an increase in CL-30B content, overall increase in total SFE can be observed due to significant improvement in acidic component. Total SFE value increased by 5.34% when CL-30B increased from 0 to 6%. This increase in total SFE is expected to improve the resistivity to moisture damage of asphaltic mixture. The ratio of acidic to basic component of SFE for asphalt binder may play an important role in deciding the bond of adhesion between asphalt and aggregate surface. With increase in acid to base energy ratio of asphalt binder, the bond of adhesion between acidic aggregate such as sandstone and granite may decrease and hence they may be susceptible to moisture damage [35]. Therefore, with increased acidic to basic energy ratio component of asphalt binder, aggregate of basic nature such as limestone and basalt may be preferred over sandstone and granite. In the present study, it was observed that acid to base surface energy component increases with an increase in CL-30B. For example, acidic to basic surface energy component ratio of virgin asphalt was found to be 0.03, which increased to 0.38 with addition of 6% CL-30B. Hence, it may be concluded that CL-30B modified binder may be least preferred in case of asphaltic mixture with acidic nature aggregates. 5.3.3. SFE component of aggregate Four different kind of aggregate types (basalt, limestone, granite and sandstone) were selected in the present study in order to capture the wide range of surface charge ranging from acidic to basic in nature. Different SFE component of aggregates selected in this study has been adopted from previously published literature (Bhasin et al. [29]) due to unavailability of instrument for measurement of SFE components of aggregates and presented in Table 3. The SFE of aggregates from literature have also been considered by many

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

Surface Free Energy Components (mJ/m2)

346

14.00 12.00

10.97

Acidic component Basic component Van der Wall component

10.43

10.00

7.99

7.16

8.00

6.25

6.87

6.18

6.00

3.99 4.00 2.00 0.00

2.66 0.35 AC-10+0%CL-30B

1.05

0.82 AC-10+2%CL30B

AC-10+4%CL30B

AC-10+6%CL30B

Sample Type

(a) Total Surface Energy (mJ/m2)

12.80

Total SFE

12.55

12.60 12.40

12.26

12.20

12.09 11.91

12.00 11.80 11.60 11.40

AC-10+0%CL-30B

AC-10+2%CL30B

AC-10+4%CL30B

AC-10+6%CL30B

Sample Type

(b) Fig. 4. Variation of (a) different SFE components and (b) total SFE for CL-30B modified asphalt binders.

Table 3 SFE components of aggregates (Bhasin et al. [29]). Aggregate type

Basalt Limestone Granite Sandstone a

SFE components (mJ/m2)a LW component (cLW)

Basic() component (c)

Acidic(+) component (c+)

Total energy (cTotal)

Acid-base part of free energy (cAB)

Acidic/basic comp (c+/c-)

52.3 44.1 48.8 58.3

164 259 412 855

0.64 2.37 0 14.6

72.8 93.6 48.84 281.75

20.49 49.55 0.00 223.45

0.004 0.009 0.000 0.017

Components were obtained by conducting Universal Sorption Device on aggregate samples.

researchers for evaluating the moisture damage potential [4,5]. It can be observed that most acidic aggregate is sandstone with acidic/basic component ratio of 0.017. The total SFE is highest in case of sandstone, whereas it is the lowest for basalt. 5.3.4. Work of adhesion in dry condition (DWdry) A higher value of DWdry is desired for a moisture resistant asphaltic mix [5,35]. It can be observed from Fig. 5(a) that DWdry increases with an increase in CL-30B content irrespective of the nature of the aggregate. The highest and lowest value of DWdry was found to be for sandstone and basaltic aggregate respectively. DWdry value increased by 41% for sandstone when CL-30B content increased from 0 to 6%. Though limestone aggregate and granite aggregate are having different SFE components, DWdry for both the cases can be observed to be similar (0, 4 and 8% difference at 2 and 4 and 6% CL-30B respectively) irrespective of CL-30B content. Considering 6 % CL-30B content, DWdry value improved by 16, 25, and 94% when aggregate type changed from basalt to limestone, granite and sandstone respectively. Hence, damage potential to

the asphaltic mixture under dry condition with CL-30B modified binder will be highest with aggregate of basaltic nature, whereas lowest with aggregate of sandstone type. 5.3.5. Work of debonding in wet condition (DWwet) A higher magnitude of DWwet indicates higher thermodynamic potential to weaken the interfacial bond between asphalt and aggregate [22]. Hence, it is desirable to have low value of DWwet which will require higher energy for breaking the interfacial bond between binder and aggregate phase. Fig. 5(b) represents the variation of DWwet for CL-30B modified asphalt binders with different types of aggregate. It can be observed that DWwet decreases with an increase in CL-30B content irrespective of nature of aggregate. For example, increase in CL-30B content from 0 to 6% resulted into reduction in DWwet by 28, 26, 27 and 21% for basalt, limestone, granite and sandstone respectively. Contrary to the DWdry, work of debonding for limestone is significantly lower than granite aggregate (35%, 34% and 33% lower at 2 and 4 and 6% CL-30B respectively). Hence it can be concluded that basaltic aggregate

347

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

180.00

Basalt

Lime stone

Granite

Sandstone

146

ΔWdry (mJ/m2)

150.00

120

116

120.00

103

90.00

61 67 64

72 72

64

69

87

76 79

94

75

60.00 30.00 0.00

AC-10+0%CL-30B

AC10+2%CL-30B

AC10+4%CL-30B

AC10+6%CL-30B

Sample type

(a) Sample type -40.00

ΔWwet (mJ/m2)

-80.00 -120.00

AC-10+0%CL-30B

AC10+2%CL-30B

-64

-60

AC10+6%CL-30B

-51

-46 -92 -106

-118

-127

-69

-79

-87

-94

AC10+4%CL-30B

-160.00

-175

-200.00 -240.00

-200

-209

-222

Basalt

Lime stone

Granite

Sandstone

-280.00

(b) Fig. 5. Variation of (a) work of adhesion and (b) work of debonding for different aggregates - asphalt binder modified with CL-30B.

1.40

Basalt

Limestone

Granite

Sandstone

1.20

1.08

Energy ratio

1.00

0.87

0.80 0.60 0.40

0

0.69

0.51 0.48

0.44 0.40

0.36 0.31

0.20

0.66

0.55

0.46

0.00

0.75

0.67 0.59

0.90

2

4

6

CL-30B Content (%) Fig. 6. Variation of energy ratio with CL-30B content for different aggregate types.

may have better performance against moisture damage followed by limestone, granite and sandstone based upon DWwet. Based on the DWdry and DWwet discussion presented above, it can be noted that CL-30B modified asphalt mixture with sandstone aggregate may perform better in the arid regions based upon work of adhesion, whereas basaltic aggregate may be preferred in the case where heavy rainfall occurs. However, in order to reach an appropriate conclusion, it is required to calculate the ER for all considered cases [5]. 5.3.6. Energy ratio (ER) Fig. 6 represents the variation of ER values for asphaltic mixture with different types of aggregates considered in the present study. It can be seen that ER value increases with an increase in CL-30B

content irrespective of the nature of aggregate, indicating improvement in moisture damage resistivity potential. The highest value of ER is observed for asphalt mixture with basaltic aggregate followed by limestone, granite and sandstone. It can be observed that asphaltic mixture with granite and sandstone type of aggregate is having similar ER value irrespective of the CL-30B content. Moisture susceptibility category for asphaltic mixture with granite and sandstone type of aggregate improved from very poor [D] to poor [C] category (refer section 3.1) when CL-30 content increased from 0 to 6%. Highest value of ER is observed for asphaltic mixture with basaltic aggregate type followed by limestone, granite and sandstone which is consistent with the work of debonding as explained in previous section. ER value increased from 0.59 to 1.08 in case of basaltic aggregate when CL-30B content increased

348

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

Table 4 Different parametersab obtained from VECD analysis.

a b

Sample code

A

B

a

sin (104)

smax (105)

C0

C1

C2

AC-10 + 0%CL-30B AC-10 + 2%CL-30B AC-10 + 4%CL-30B AC-10 + 6%CL-30B

114,346 248,775 459,442 1,448,776

3.736 3.808 4.182 4.803

1.870 1.904 2.092 2.401

1.05 1.14 1.41 2.23

3.14 3.48 4.0 5.83

1 1 1 1

0.087 0.093 0.110 0.149

0.450 0.432 0.399 0.353

C1, C2 are curve fitting coefficients, sin and smax are initial and peak shear stress in Pa. Represents average values.

Integrity parameter (C)

1

AC-10+0%CL-30B AC-10+4%CL-30B

0.8

AC-10+2%CL-30B AC-10+6%CL-30B

0.6

0.4

0.2

0

0

50

100

150

200

250

300

350

Damage Intensity (D) Fig. 7. Integrity parameter versus damage intensity for CL-30B modified asphalt binders.

load cycle to failure (Nf)

1.00E+08 AC-10+0%CL-30B

AC-10+2%CL-30B

AC-10+4%CL-30B

AC-10+6%CL-30B

1.00E+06

1.00E+04

1.00E+02

Nf_0%CL-30B = 114346(%ε)-3.736 Nf_2%CL-30B = 248775(%ε)-3.808 Nf_4%CL-30B = 459442(%ε)-4.182 Nf_6%CL-30B = 1448776(%ε)-4.803

1.00E+00

0.5

Strain Level (%)

5

Fig. 8. Load cycle to failure versus strain level for CL-30B modified asphalt binders.

from 0 to 6% which improves moisture damage resistivity from poor [C] to fair [B] category. Similar improvement for asphaltic mixture with aggregate of limestone nature can also be observed. Finally, it can be observed that moisture susceptibility was in very poor [D] category without addition of CL-30B (except basalt which falls in poor [C] category) which improved their category to the next class with addition of CL-30B in all the cases. Overall, it can be concluded that, in the case of CL-30B modified asphaltic mixture with aggregate of basaltic origin will have best performance against moisture damage among all considered cases. 5.4. Fatigue performance 5.4.1. Damage intensity (D) versus Integrity parameter (C) Table 4 represents the various parameter obtained from VECD analysis such as undamaged material property (a), initial and maximum shear stress values (sin and smax), and different components of integrity parameter Vs damage intensity equation (Eq. (6)). sin and smax value increased by 2.12 and 1.85 times respectively when

CL-30B content changed from 0 to 6% due to increase in stiffness value. Value of C equal to 1 represents the highest integrity with no damage, whereas C equal to zero represents the complete failure to material. Fig. 7 shows that the rate of decrease of integrity level to asphalt binder is increasing with increase in CL-30B content. It can also be observed that at any particular level of integrity, the damage to asphalt is highest in case of 0% CL-30B content and lowest with 6% CL-30B content. 5.4.2. Fatigue life Table 4 also shows the value of fatigue life equation coefficients obtained LAS test data. It is to be noted that factor ‘‘A” is obtained after carrying out damage analysis of asphalt binder from amplitude sweep test, whereas ‘‘B” represents the undamaged material property. Eq. (8) shows that higher the value of ‘‘A”, higher will be the fatigue life of the asphalt binder. Table 4 shows increase in ‘‘A” value with an increase in CL-30B content which ultimately increases the fatigue life of asphalt binder. Micaelo et al. [16] observed similar trend of increase in fatigue life equation parameter for SBS modified

349

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

1.00E+08

G* (Pa)

1.00E+07

1.00E+06

1.00E+05

Points of LVE Limits AC-10+0%CL-30B AC-10+2%CL-30B AC-10+4%CL-30B AC-10+6%CL-30B

1.00E+04 1.00E-01

1.00E+00

1.00E+01

1.00E+02

Strain (%) Fig. 9. Complex modulus versus strain obtained from amplitude sweep test.

Table 5 Limiting LVE values, limiting complex and shear modulus values from amplitude strain sweep test at 25 °C. Sample code

Limiting LVE strain (%)

Limiting complex modulus value (kPa)

Limiting shear stress value (kPa)

AC-10 + 0%CL-30B AC-10 + 2%CL-30B AC-10 + 4%CL-30B AC-10 + 6%CL-30B

1.35 1.37 1.18 1.02

10,027 10,830 13,442 21,185

133 145 157 215

and neat binder. Considering the strain level of 1%, fatigue life value increased from 114,346 to 1,448,776 with corresponding increase of CL-30B content from 0 to 6%. Such improvement in fatigue life of could be the result of nanoclay cluster formation in asphalt matrix which may prevent formation of crack [17]. It is also important to notice that with increase in strain level values, the difference in their fatigue life value is diminishing (see Fig. 8). 5.4.3. Linear visco-elastic range In this study, an attempt has been made to check whether the addition of CL-30B has influences on Linear Visco-Elastic (LVE) limit of asphalt binder or not. Fig. 9 shows the variation of complex modulus value against applied amplitude strain values obtained through amplitude strain sweep test. As per the criteria set by researchers, a point corresponding to a decrease in stiffness value by 5% was assumed as a LVE limit point [36,37]. The LVE limiting values and the corresponding complex modulus and shear stress values are reported in Table 5. It can be observed that LVE limit of asphalt binder decreases with an increase in CL-30B. The similar observation was reported by Santagata et al. [38] for Carbon Nano Tube (CNT) modified asphalt binder. 6. Conclusions This present study examines rutting, moisture damage and fatigue performance of CL-30B modified asphalt binder. Superpave rutting parameter was used for investigating the rutting performance while the SFE approach and the LAS test were selected for evaluating moisture damage and fatigue performance respectively. The following conclusions can be drawn from the results and discussion presented above:
Improvement in rutting performance of asphalt binder was observed with addition of CL-30B. The PG grade of the control binder increases by 1 PG and 2 PG with addition of 4% and 6% of CL-30B; however, no PG grade change was noticed for 2% CL-30B modified binder.


The SFE component measured from Wilhelmy plate method shows that acidic component of asphalt binder increases, while basic and Vander wall components decrease with addition of CL-30B. The total SFE of asphalt binder increases with an increase in CL-30B content.
Work of adhesion increases with addition of CL-30B, indicating CL-30B helps in improving aggregate-asphalt binder bond in dry condition. The highest value of work of adhesion was observed for sandstone aggregate followed by granite, limestone and basalt aggregates.
Work of debonding decreases with an increase in CL-30B content irrespective of aggregate type considered which is desirable for improvement in moisture damage resistivity. Lowest value of work of debonding was observed for basaltic aggregate.
Increase in ER value was observed with addition of CL-30B to asphalt binder irrespective of type of aggregate considered. Based on ER value, it may be concluded that addition of CL30B to asphalt binder may result in a moisture resistant mix. The ER value of basaltic aggregates was found to be in the highest followed by limestone, sandstone and lowest for granite type of aggregate. This shows basalt aggregates may have a better moisture resistant among all the consider aggregates types followed by limestone, sandstone and lowest for granite type.
Based on the LAS test, it was found that number of load cycle to failure increases with an increase in CL-30B content. Results show improvement in fatigue life of asphalt binder with addition of CL-30B. However, this inference cannot be generalized and needs further detailed investigations in this direction for reaching at appropriate conclusion in this regard.
The LVE range of asphalt binder decreases with addition of CL30B, indicating a stiffer binder with addition of nanoclay. It is expected the data generated in the present study would be helpful to understand behaviour of nanoclay modified asphalt binders.

7. Limitations and recommendation The current study presents conclusions based on the limited laboratory investigations conducted on nanoclay modified asphalt binder. The behaviour of nanoclay modified asphalt binder may change with type, source and chemical composition of control binder used, thus a future study needs to be conducted in this direction. In addition, rutting resistance of nanoclay modified mixes should be evaluated in the laboratory. The moisture sensitive in the present research study has been investigated using SFE components of asphalt binder and aggregates, however, an appropriate laboratory study should be further conducted on asphalt mixes

350

P.K. Ashish et al. / Construction and Building Materials 113 (2016) 341–350

to validate the result obtained through this approach. Further, LAS test in the present study indicated increase in fatigue life, although stiffness of binder increases with addition of nanoclay, thus it is recommended that a future study be conducted to evaluate fatigue life of asphalt mixes modified with nanoclay to reach at appropriate conclusion regarding fatigue life. Acknowledgements The authors would like to acknowledge Department of Earth Science, IIT-Bombay for providing facility for conducting XRD studies on various samples used in the present study. References [1] Z. You, J. Mills-Beale, J.M. Foley, S. Roy, G.M. Odegard, Q. Dai, S.W. Goh, Nanoclay modified asphalt materials: preparation and characterization, Constr. Build. Mater. 25 (2) (2011) 1072–1078. [2] N.I.M. Yusoff, A.A.S. Breem, H.N. Alattug, A. Hamim, The effects of moisture susceptibility and ageing conditions on nano-silica/polymer-modified asphalt mixtures, Constr. Build. Mater. 72 (15) (2014) 139–147. [3] G.H. Hamedi, F.M. Nejad, K. Oveisi, Estimating the moisture damage of asphalt mixture modified with nano zinc oxide, Mater. Struct. (2015) 1–10, http://dx. doi.org/10.1617/s11527-015-0566-x. [4] A. Diab, Z. You, Effects of regular-sized and nanosized hydrated lime on binder rheology and surface free energy of adhesion of foamed warm mix asphalt, J. Mater. Civ. Eng. 27 (9) (2014) 1–7. [5] Z. Hossain, M. Zaman, T. Hawa, M.C. Saha, Evaluation of moisture susceptibility of nanoclay-modified asphalt binders through the surface science approach, J. Mater. Civ. Eng. 27 (10) (2014) 1–9. [6] A.Z. Shahabadi, A. Shokuhfar, S.E. Nejad, Preparation and rheological characterization of asphalt binders reinforced with layered silicate nanoparticles, Constr. Build. Mater. 24 (7) (2010) 1239–1244. [7] G. Liu, M. van de Ven, S. Wu, J. Yu, A. Molenaar, Influence of organomontmorillonites on fatigue properties of bitumen and mortar, Constr. Build. Mater. 33 (12) (2011) 1574–1582. [8] H. Yao, H. Yao, Z. You, L. Li, S. Goh, J. Mills-Beale, X. Shi, D. Wingard, Evaluation of asphalt blended with low percentage of carbon micro-fiber and nanoclay, J. Test. Eval. 41 (2) (2013) 1–11. [9] S Wu, J Wang, L Jiesheng, Preparation and fatigue property of nanoclay modified asphalt binder, Mechanic Automation and Control Engineering (MACE), IEEE, 2010, pp. 1595–1598. [10] J.Y. Yu, P.C. Feng, H.L. Zhang, S.P. Wu, Effect of organo-montmorillonite on aging properties of asphalt, Constr. Build. Mater. 23 (7) (2009) 2636–2640. [11] S.G. Jahromi, N.A. Ahmadi, S.M. Mortazavi, S. Vossough, Rutting and fatigue behavior of nanoclay modified bitumen, Int. J. Sci. Technol. Trans. Civ. Eng. 35 (C2) (2011) 277–281. [12] D.B. Ghile, Effects of nanoclay modification on rheology of bitumen and on performance of asphalt mixtures (MS thesis), Delft University of Technology, Delft, 2006. [13] J.S. Ghaffarpour, B. Andalibizade, A. Khodaii, Mechanical behavior of nanoclay modified asphalt mixtures, J. Test. Eval. 38 (5) (2010) 1–9. [14] C. Hintz, R.A. Velasquez, C.M. Johnson, H.U. Bahia, Modification and validation of the linear amplitude sweep test for binder fatigue specification, J. Transp. Res. Board 2207 (2011) 99–106. [15] C.M. Johnson, H.U. Bahia, Evaluation of an accelerated procedure for fatigue characterization of asphalt binders, Road Mater. Pavement Des., submitted for publication, Available at: uwmarc.wisc.edu/files/linearamplitudesweep/ RMPD10_LAS_CMJ_HB-100321.pdf (access date: 05-09-2015). [16] R. Micaelo, A. Pereira, L. Quaresma, M.T. Cidade, Fatigue resistance of asphalt binders: assessment of the analysis methods in strain-controlled tests, Constr. Build. Mater. 98 (15) (2015) 703–712.

[17] A. Kavussi, P. Barghabany, Investigating fatigue behavior of nanoclay and nano hydrated lime modified bitumen using LAS test, J. Mater. Civ. Eng. 28 (3) (2015) 889–897. [18] B. Golestani, B.H. Nam, F.M. Nejad, S. Fallah, Nanoclay application to asphalt concrete: characterization of polymer and linear nanocomposite-modified asphalt binder and mixture, Constr. Build. Mater. 91 (30) (2015) 32–38. [19] D.N. Little, A. Bhasin, Using surface energy measurements to select materials for asphalt pavement, Transp. Res. Board NCHRP Project (2006) 9–37. [20] M. Arabani, H. Roshani, G.H. Hamedi, Estimating moisture sensitivity of warm mix asphalt modified with zycosoil as an antistrip agent using surface free energy method, J. Mater. Civ. Eng. 24 (7) (2012) 889–897. [21] A. Bhasin, D. Little, Characterizing Surface Properties of Aggregates used in Hot Mix Asphalt, Rep. Int. Centre for Aggregates Research (ICAR)/505-2, Aggregates Foundation for Technology, Research, 2006. [22] A Bhasin, J Howson, E Masad, DN Little, RL Lytton, Effect of modification processes on bond energy of asphalt binders, Transportation Research Board 86th Annual Meeting, Washington, DC, 2007, pp. 1–14. [23] A. Bhasin, J.W. Button, A. Chowdhury, Evaluation of Selected Laboratory Procedures and Development of Databases for HMA. Rep. No., Texas Transportation Institute, Texas A&M Univ, Federal Highway Administration (FHWA)/Texas (TX), 2005. [24] C.J. Van Oss, M.K. Chaudhury, R.J. Good, Interfacial Lifshitz-van der Waals and polar interactions in macroscopic systems, Chem. Rev. 88 (1988) 927–941. [25] C.M. Johnson, Estimating asphalt binder fatigue resistance using an accelerated test method (Ph.D. thesis), Civil & Environmental Engineering, University of Wisconsin, 2010. [26] M.E. Kutay, N. Gibson, J. Youtcheff, Conventional and viscoelastic continuum damage (VECD) based fatigue analysis of polymer modified asphalt pavements, J. Assoc. Asphalt Paving Technol. (AAPT) 77 (2008) 395–434. [27] AASHTO: TP 101, Estimating Damage Tolerance of Asphalt Binders Using the Linear Amplitude Sweep, American Association of State Highway and Transportation Officials, 2014, pp. 1–7. [28] H. Bahia, H.A. Tabatabaee, T. Mandal, A. Faheem, Field evaluation of Wisconsin modified binder selection guidelines. Phase II, Wisconsin Highway Research Program WisDOT ID No. 0092-13-02, 2013. [29] A. Bhasin, D. Little, Characterization of aggregate surface energy using the universal sorption device, J. Mater. Civ. Eng. 19 (8) (2007) 634–641. [30] ASTM: D 1754, Standard Test Method for Effects of Heat and Air on Asphaltic Materials (Thin-Film Oven Test), American Society for Testing and Material, 2014, pp. 1–6. [31] ASTM:D 6521, Standard Practice for Accelerated Aging of Asphalt Binder Using a Pressurized Aging Vessel (PAV), American Society for Testing and Material, 2013, pp. 1–6. [32] ASTM:D 7175, Standard Test Method for Determining the Rheological Properties of Asphalt Binder Using a Dynamic Shear Rheometer, American Society for Testing and Material, 2015, pp. 1–16. [33] A.W. Hefer, A. Bhasin, D.N. Little, Bitumen surface energy characterization using a contact angle approach, J. Mater. Civ. Eng. 18 (6) (2006) 759–767. [34] U. Yuan, T.R. Lee, Surface science techniques, Contact Angle and Wetting Properties, Department of Chemistry, University of Houston, 2013, pp. 3–34, http://dx.doi.org/10.1007/978-3-642-34243-1_1. [35] M. Arabani, G.H. Hamedi, Using the surface free energy method to evaluate the effects of polymeric aggregate treatment on moisture damage in hot mix asphalt, J. Mater. Civ. Eng. 23 (6) (2011) 802–818. [36] JC Petersen, RE Robertson, JF Branthaver, DA Anderson, DW Christiansen, HU Bahia, Binder characterization and evaluation, Strategic Highway Research Program, SHRP-A-367, vol. 1, 1994. [37] D.A. Anderson, D.W. Christensen, H.U. Bahia, R. Dongre, M.G. Sharma, C.E. Antle, J. Button, Binder characterization and evaluation, Strategic Highway Research Program, SHRP-A-369, vol. 3, 1994. [38] E. Santagata, O. Baglieri, L. Tsantilis, G. Hiappinelli, Fatigue properties of bituminous binders reinforced with carbon nanotubes, Int. J. Pavement Eng. 16 (1) (2015) 80–90.