Relating Dynamic Shear Modulus (G*) of Performance Grade (PG) Asphalt Binders to Nanoindentation Stiffness (E)

Transcription

1 Final Report September 2016 Relating Dynamic Shear Modulus (G*) of Performance Grade (PG) Asphalt Binders to Nanoindentation Stiffness (E) SOLARIS Consortium, Tier 1 University Transportation Center Center for Advanced Transportation Education and Research Department of Civil and Environmental Engineering University of Nevada, Reno Reno, NV Rafiqul A. Tarefder, Ph.D., P.E. Professor of Civil Engineering and Regents Lecturer Associate Director, SPTC, Associate Director, SOLARIS, University of New Mexico; Albuquerque, NM

2 DISCLAIMER: The contents of this report reflect the views of the authors, who are responsible for the facts and accuracy of the information presented herein. This document is disseminated under the sponsorship of the U.S. Department of Transportation s University Transportation Centers Program, in the interest of information exchange. The U.S. Government assumes no liability for the contents or use thereof. 2

3 SUMMARY PAGE 1. Report No. 2. Recipient s Catalog No. 3. Title and Subtitle Relating Dynamic Shear Modulus (G*) of Performance Grade (PG) Asphalt Binders to Nanoindentation Stiffness (E) 5. Author(s): Hasan M. Faisal, Umme Amina Mannan and Rafiqul A. Tarefder 7. Performing Organization Name and Address University of New Mexico Department of Civil Engineering MSC University of New Mexico Albuquerque, NM Sponsoring Agency Name and Address SOLARIS Institution, University Transportation Center and New Mexico Department of Transportation (NMDOT) 4. Report Date September 31, Performing Organization Report No. 8. Performing Organization Code 9. Contract/Grant No. 11. Type of Report and Period Covered Final Report 12. Sponsoring Agency Code 13. Supplementary Notes The research project is funded by SOLARIS in cooperation with the NMDOT 14. Abstract Traditionally, asphalt binder is characterized by a dynamic shear rheometer, which applies a shear load on a bulk volume of liquid asphalt binder to determine shear modulus (G*). Recently, a nanoindentation test can characterize an asphalt binder film in the form of coating around roadway aggregates, which is more practical. In a nanoindentation test, a sharp tip is used to indent an asphalt film while residing on an aggregate surface to determine nanoindentation modulus (E). This study evaluates whether there is a relation between G* and E. For both tests, replicate samples were conditioned in three ways: unaging, rolling thin film oven aging, and pressure vessel aging. Results show that the E-value is approximately 2 to 6 times larger than the G*-value based on all samples/conditioning. 15. Key Words Nanoindentation, Asphalt, Complex, Modulus, Rheology 17. Security Classi. of the Report None 18. Security Classi. of this page None 16. Distribution Statement No restrictions. 19. Number of 20. Price Pages N/A 31 3

4 Table of Contents Introduction... 6 Objectives... 7 Task/Objective 1: Asphalt Binder collection... 7 Task/Objective 2: Conditioning and DSR Testing of PG Binders... 8 Conditioning... 8 Dynamic Shear Rheometer (DSR) Testing... 8 Results and Discussion... 9 Aging profile of binders Task/Objective 3: Sample Preparation and Nanoindentation Testing of PG Binders Nanoindentation Oliver-Pharr Method Materials and Sample Preparation Nanoindentation Test Results and Discussion Task/Objective 4: Relating G* to E Conclusions Recommendations References

5 List of Figures Figure 1: Laboratory Aging Test apparatus 7 Figure 2: Dynamic Shear Rheometer 8 Figure 3: Master curve of PG binder (reference temperature 21.1 C (70 F)) 9 Figure 4: Master curve of PG binder (reference temperature 21.1 C (70 F)) 10 Figure 5: Master curve of PG binder (reference temperature 21.1 C (70 F)) 11 Figure 6: Master Curve of PG binder (reference temperature 21.1 C (70 F)) 11 Figure 7: Master Curve of PG binder (reference temperature 21.1 C (70 F)) 12 Figure 8: Master Curve of PG binder (reference temperature 21.1 C (70 F 13 Figure 9: Master curve of PG binder (reference temperature 21.1 C (70 F)) 14 Figure 10: Master curve of PG binder (reference temperature 21.1 C (70 F)) 15 Figure 11: effect of aging on G* value at 10 rad/sec frequency 16 Figure 12: effect of aging on phase angle (δ) value at 10 rad/sec frequency 17 Figure 13: Schematic of the Indentation Test 18 Figure 14: Asphalt Binder Sample for Nanoindentation 21 Figure 15: Nanoindentation Test on Asphalt Binder 22 Figure 16: Load Displacement Curve 23 Figure 17: Modulus for different PG binder grades and three different aging conditions 23 Figure 18: Hardness for different PG binder grades and three different aging conditions 24 Figure 19: Comparison of E and G* 25 Figure 20: Pearson Correlation between E and G* of Asphalt Binder 26 Figure 21: Non Linear Correlation between E and G* of Asphalt Binder 27 List of Tables Table 1: Asphalt binders collected 6 5

6 Introduction Asphalt concrete s dynamic modulus (E*) is one of the key input parameters for structural design of flexible pavement according to the Mechanistic Empirical Pavement Design Guide (MEPDG, currently packaged as DARWin-ME software). Currently, asphalt binder shear modulus G*-based Witczak equation is used in MEPDG Level 2 analysis to estimate E* of asphalt concrete from Hot Mix Asphalt (HMA) mix volumetrics and binder information. Specifically, asphalt binder s dynamic shear modulus (G*) and phase angle (δ) are used in G*-based Witczak equation to predict E* mastercurve at Level 2 MEPDG analysis. While Asphalt Concrete s (AC) E* testing is conducted using an axial compression (or tension), asphalt binder s G* is conducted in shear mode. Because, over the years, test methods developed and performed to characterize asphalt binders have been limited to a few rheological tests. Moreover, the existing tests cannot be performed on asphalt binder film while they are integral parts of AC. Rather those tests are performed on bulk liquid asphalt separately. Nanoindentation has created an opportunity to determine the stiffness (E) of asphalt binder while they are parts of AC and under compression loading. Aging/oxidative aging of asphalt is one of the major concerns in the asphalt research field, as aging can lead to the development of distresses like fatigue or thermal cracking in asphalt pavement (1-2). Aging in asphalt leads to stiffening of the material as well as it changes in its chemical composition. It is well known when asphalt ages the nonpolar molecule part decreases and oxidative aged polar part increases (3). The increased amounts of polar molecules change the nature of the asphalt and lead to increased stiffness and slower stressrelaxation rates. Understanding increased stiffness and slower stress-relaxation is important for asphalt researchers, in other words, understanding of the response of stiffening of asphalt can lead to attractive design parameters for pavement. In a nanoindentation test, a sharp indenter is loaded to indent an asphalt sample surface and the displacement of the indenter is measured as a function of the load. Load, displacement and time data are recorded when the indenter indents and retracts. Modulus of elasticity, hardness of a material are determined from the load-displacement data. Though the properties of hard materials such as metals and polymeric composites are commonly determined by nanoindenter; asphalt binder (film) has not tested because they are soft. This study takes the challenge of indenting soft, viscous materials such as asphalt binder subjected to different aging. To date, researchers have performed various tests on mastic and asphalt binder at micro scale to understand the macro scale behavior of AC (1-9). However, test methods developed and performed on binders, to this day, are mostly rheological tests. Very few studies have been conducted on the compression stiffness and hardness of binder, rather than shear stiffness (10-15). The existing tests used in the asphalt area cannot be performed on thin film asphalt binder while thin film asphalt binder coats plays major role in asphalt concrete. Rather; the tests developed by Superpave effort are performed on the bulk asphalt binder. Nanoindentation has created an opportunity to characterize thin film asphalt binder for different aging condition. The introduction of nanoindentation in the field of asphalt researchers is rather limited (10-15). Asphalt is known to be a viscoelastic material that exhibits creep behavior. Tarefder et al. developed a range of indentation derived modulus and hardness values of AC and 6

7 asphalt binder (10). Jager et al. studies the thermal effects on the mechanical properties of the asphalt binder (8). Tarefder and Faisal has done nanoindentation of AC, however the study only covered only two phases of AC, mastic and aggregate also limited number of indentations and 5 years of oven aging were done in that study (12). In another study Tarefder et al. showed the effect of aging of mica fines mixed asphalt binder, which concluded at higher percentage of fines aged mastic becomes soft and brittle (19). Faisal et al. showed the effect of moisture damage on different phases of AC using nanoindentation (20). To this end, this study proposes to conduct dynamic shear modulus (G*) and nanoindentation stiffness (E) testing of selected Performance Grade (PG) asphalt binders of New Mexico sources and examine possible correlation between G* and E. It is expected that asphalt binder s E will a better estimator of asphalt concrete s E* for Level 2 MEPDG analysis. The proposed study will increase reliability in the design of pavement infrastructure in the transportation industry sectors, and provide the knowledge base to synthesize binders using nanotechnology for advanced asphalt pavements. Therefore, the proposed project will develop new cost-saving processing capabilities based on asphalt binder with better known rheological characteristics and improved durability against wear and tear. The outcome of this study may generate nano-engineering concept to characterize new asphalt resulting in drastically improved asphalt characteristics including, but not limited to, superior aging and reduced moisture-induced damages. Objectives This study was conducted to provide a comprehensive evaluation of the material characteristics and design of thin asphalt binder in nano and micro scale. The major tasks carried out in this research were: Collect and conditioning eight different asphalt binders. Short term and long term aging of all the binders were done using Rolling Thin Film Oven (RTFO) and Pressure Aging Vessel (PAV). Rheological material characterization of asphalt binder using dynamic shear modulus. Nanomechanical characterization of thin film asphalt binder using nanoindentation testing. Task/Objective 1: Asphalt Binder collection A total of eight performance grade (PG) binders available in New Mexico were collected in five one-gallon buckets per PG binder. Table 1: Asphalt binders collected PG PG PG PG PG PG PG PG

8 Some performance graded asphalt binder such as PG 58-34, PG 64-34, PG are not used in New Mexico. These binders could not be collected. Task/Objective 2: Conditioning and DSR Testing of PG Binders Conditioning One PG binders were subjected to three conditioning (1) unaged/ original binder and (2) RTFO (Rolling Thin Film Oven) and (3) PAV (Pressure Aging Vessel) aged before DSR testing. The Rolling Thin Film Oven (RTFO) Test simulates the short-term aging of asphalt binders that occurs during the hot-mixing process. Asphalt binder was exposed to elevated temperature to simulate manufacturing and placement aging according to AASHTO T (2009) test procedure. Unaged asphalt binder was taken in cylindrical glass bottles and placed in a rotating carriage within an oven which rotates within the oven at 325 F (163 C) temperature for 85 minutes. Figure 1(a) shows the RTFO apparatus. After the RTFO aging, 50gm sample was placed in the PAV plate and then aged at 100 C and 2.10 MPa for 20 hours according to the AASHTO R Figure 1(b) shows the PAV apparatus. a) Rolling Thin-Film Oven (RTFO) b) Rolling Thin-Film Oven (PAV) Figure 1: Laboratory Aging Test apparatus Dynamic Shear Rheometer (DSR) Testing Dynamic Shear Rheometer (DSR) was conducted on the conditioned binders. The test measures the dynamic shear modulus or complex modulus, G* and phase angle (δ) of asphalt binders under a continuous sinusoidal loading using a dynamic shear rheometer and parallel plate test geometry. Figure 2 shows the DSR test device. 8

9 Figure 2: Dynamic Shear Rheometer The G* and δ were measured at seven different temperatures (40, 55, 70, 85, 100, 115 and 130 F) and at different loading frequency. During the test, the asphalt binder sample was subjected to shear stress at frequencies ranging from 1 rad/s to 100 rad/s with 5 measuring points per decade in log scale. At each temperature, G* values were tested at 11 different frequencies (1, 1.58, 2.51, 3.98, 6.31, 10, 15.8, 25.1, 39.8, 63.1, and 100 rad/sec). At each temperature the binders stiffness (G*) and phase angle (δ) values at different frequency were recorded. A thicker sample (2 mm) with a smaller diameter plate (8 mm) was used so that the phase angle (δ) becomes measurable. All the tests were conducted following the AASHT0 T (2009) test protocol within the linear viscoelastic range under straincontrolled mode. Three replicate samples were tested and the average of the three tests was used to develop the master curve. Results and Discussion PG binder Figure 3 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. In master curve, higher frequency represents low temperature and lower frequency represents high temperature. It can be observed that the G* value of the asphaltic material increases with aging. PAV aged binder has higher G* value over the whole frequency range. At high frequency or low temperature, the effect of aging is less than the low frequency or high temperature. Therefore, it can be said that short-term and long-term aging mostly effect the high temperature performance of binder. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 9

10 1.0E+8 Dynamic shear modulus, G* (Pa) 1.0E+7 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG58-22_Unaged PG58-22_RTFO PG58-22_PAV 1.0E+0 1.0E-4 1.0E-3 1.0E-2 1.0E-1 1.0E+0 1.0E+1 1.0E+2 1.0E+3 1.0E+4 Reduced Frequency, f r (rad/s) Figure 3: Master curve of PG binder (reference temperature 21.1 C (70 F)) PG binder Figure 4 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. Similar to other binder, the G* value of the asphaltic material increases with aging. PAV aged binder has higher G* value over the whole frequency range. At high frequency or low temperature, the effect of aging is less than the low frequency or high temperature. Therefore, it can be said that short-term and long-term aging mostly effect the high temperature performance of binder. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 10

11 1.0E+8 1.0E+7 Dynamic shear modulus, G* (Pa) 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG58-28_Unaged PG58-28_RTFO PG58-28_PAV 1.0E+0 1.0E-4 1.0E-2 1.0E+0 1.0E+2 1.0E+4 Reduced Frequency, f r (rad/s) Figure 4: Master curve of PG binder (reference temperature 21.1 C (70 F)) PG binder Figure 5 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. Similar to other binder, the G* value of the asphaltic material increases with aging. PAV aged binder has higher G* value over the whole frequency range. At high frequency or low temperature, the effect of aging is less than the low frequency or high temperature. Therefore, it can be said that short-term and long-term aging mostly effect the high temperature performance of binder. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 11

12 1.0E+8 1.0E+7 Dynamic shear modulus, G* (Pa) 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG64-22_Unaged PG64-22_RTFO PG64-22_PAV 1.0E+0 1.0E-4 1.0E-3 1.0E-2 1.0E-1 1.0E+0 1.0E+1 1.0E+2 1.0E+3 1.0E+4 Reduced Frequency, f r (rad/s) Figure 5: Master curve of PG binder (reference temperature 21.1 C (70 F)) PG binder Figure 6 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. Similar to other binder, the G* value of the asphaltic material increases with aging. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 1.0E+8 1.0E+7 Dynamic shear modulus, G* (Pa) 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG64-28_Unaged PG64-28_RTFO PG64-28_PAV 1.0E+0 1.0E-4 1.0E-2 1.0E+0 1.0E+2 1.0E+4 Reduced Frequency, f r (rad/s) Figure 6: Master Curve of PG binder (reference temperature 21.1 C (70 F)) 12

13 PG binder Figure 7 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. Similar to other binder, the G* value of the asphaltic material increases with aging. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 1.0E+8 1.0E+7 Dynamic shear modulus, G* (Pa) 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG70-22_Unaged PG70-22_RTFO PG70-22_PAV 1.0E+0 1.0E-4 1.0E-2 1.0E+0 1.0E+2 1.0E+4 Reduced Frequency, f r (rad/s) Figure 7: Master Curve of PG binder (reference temperature 21.1 C (70 F)) PG binder Figure 8 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. Similar to other binder, the G* value of the asphaltic material increases with aging. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 13

14 1.0E+8 1.0E+7 Dynamic shear modulus, G* (Pa) 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG70-28_Unaged PG70-28_RTFO PG70-28_PAV 1.0E+0 1.0E-4 1.0E-2 1.0E+0 1.0E+2 1.0E+4 Reduced Frequency, f r (rad/s) Figure 8: Master Curve of PG binder (reference temperature 21.1 C (70 F)) PG binder Figure 9 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. Similar to other binder, the G* value of the asphaltic material increases with aging. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 14

15 1.0E+8 Dynamic shear modulus, G* (Pa) 1.0E+7 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG76-22_Unaged PG76-22_RTFO PG76-22_PAV 1.0E+0 1.0E-4 1.0E-3 1.0E-2 1.0E-1 1.0E+0 1.0E+1 1.0E+2 1.0E+3 1.0E+4 Reduced Frequency, f r (rad/s) Figure 9: Master curve of PG binder (reference temperature 21.1 C (70 F)) PG binder Figure 10 shows the master curve of shear modulus at 21.1 C (70 F) temperatures for unaged, RTFO and PAV aged PG binder. Similar to other binder, the G* value of the asphaltic material increases with aging. Also the RTFO and PAV aged binder has higher G* than the unaged binder at lower frequency, which means that aging of binder will provide better rutting resistance (Less permanent deformation). 15

16 1.0E+8 1.0E+7 1.0E+6 1.0E+5 1.0E+4 1.0E+3 1.0E+2 1.0E+1 PG76-28_Unaged PG76-28_RTFO PG70-28_PAV 1.0E+0 1.0E-4 1.0E-3 1.0E-2 1.0E-1 1.0E+0 1.0E+1 1.0E+2 1.0E+3 1.0E+4 1.0E+5 Figure 10: Master curve of PG binder (reference temperature 21.1 C (70 F)) Aging profile of binders In previous section the master curve for all the binders is presented for unaged, RTFO aged and PAV aged binders. In this section, the effect of aging on G* and phase angle (δ) are shown. In figure 11, all the G* values are shown at 10 rad/sec loading frequencies and at 21.1 C (70 F). For PG binder, the G* values are , and KPa for unaged, RTFO aged and PAV aged. With aging the value is increasing. Form unaged to RTFO the G* increase very little however form RTFO to PAV aging G* increases tremendously. Also PG xx-28 binders showed higher increase in G* values due to PAV aging than the PG xx-22 binders. 16

17 7000 Unaged RTFO PAV G* (KPa) Binder Type Figure 11: effect of aging on G* value at 10 rad/sec frequency. In figure 12, all the phase angle (δ) values are shown at 10 rad/sec loading frequencies and at 21.1 C (70 F). For PG binder, the δ values are 59.10, and for unaged, RTFO aged and PAV aged. With aging the value is decreasing. Softer PG binders has more decrease in the δ value than the harder PG binders. 17

18 80 70 Unaged RTFO PAV 60 Phase Angle ( ) Binder Type Figure 12: effect of aging on phase angle (δ) value at 10 rad/sec frequency. Task/Objective 3: Sample Preparation and Nanoindentation Testing of PG Binders Nanoindentation In a nanoindentation test, an indenter tip of a known modulus of elasticity and geometry is loaded to penetrate a sample surface and then unloaded. Modulus of elasticity (E) of the sample is determined from the load-displacement data. The area of contact at full load is determined from the measured depth of penetration and the known geometry of the indenter tip. Sample hardness (H) is calculated by dividing the maximum load by the contact area. A typical load displacement curve is shown in Fig. 13(a). A sitting load is typically applied initially to facilitate a contact between the tip and sample surface. Next, the load is increased gradually from point a to b. The tip is unloaded at the maximum load point b. The unloading path is assumed to be elastic for most of the elastoplastic material. The unloading curve does not come back to point a due to plastic deformation in elastoplastic materials. The slope of the unloading curve at point b is usually equal to the slope of the loading curve at point a. 18

19 a) Load Displacement Curve (b) Indentation Depth Figure 13: Schematic of the Indentation Test. Figure 13(b) shows the surface profile as a function of the penetration depth during loading and unloading. Here, hmax is the total depth of indentation at a maximum load, hp is the total depth of indentation that is unrecovered, hs is the depth of the surface at the perimeter of 19

20 the indenter contact and hc is the vertical depth along which the contact is made between the indenter and the sample. Therefore, h = h max (1) c h s The depth of impression that is recovered is, h = h max (2) e h p Oliver-Pharr Method The Oliver and Pharr method is based on the elastic contact between a rigid sphere (tip) and a flat surface (sample) (20). If the indentation load P penetration depth h is recorded as the load displacement curve, the reduced elastic modulus E* can be found from the load displacement curve. However, the equation also relates to the indentation radius. The elastic modulus of the indented sample can be inferred from following equation, where indenter radius R and contact radius a: dp = 2 E * Rh = 2E * a dh The projected area at the maximum load can be defined as: A =πa 2. Therefore, dp 2 S = = E * dh π where S is the unloading stiffness or slope of the unloading curve; A (3) (4) π E* = ( S) 2 A (5) How to Find S Oliver and Pharr used a power law function to fit the unloading path of the loaddisplacement curve (23). The power law function used by Oliver-Pharr is shown in Eq. (6): P = α( h ) where h is depth of penetration, hf is plastic depth, α and m are curve fitting parameters related to tip geometry. m is equal to 1 for flat ended cylindrical tip, m is 1.5 for spherical h f m (6) 20

21 tip, and m is 2 for conical tip (Berkovich tip). Slope is measured by differentiation in the above Eq. (6) at onset of unloading. How to Find A Oliver and Pharr defined the projected area A as a function of hc defined in Eq. (1). Oliver and Pharr extrapolated the tangent line to the unloading curve at the maximum loading point down to zero loads. This yields an intercept value for depth which estimates the hs by: h s P = ε S max (7) Therefore, h c = h max P ε S where εε is a geometric constant. ε = 0.72 for conical tip, ε = 0.75 for Berkovich tip, and ε = 0.72 for spherical tip. The project area is measured by: A 2 = π = a π max ( Rhc) where R is known and hc is calculated using the above Eq. (9). How to Find E Timoshenko and Goodier found the reduced elastic modulus, E* is related to the modulus of the indenter and the specimen and given by (Oliver and Pharr 1992): ν 1 ν = + E * E E where E is Young s modulus of the material, ν is Poisson s ratio of the material, Ei is Young s modulus of the indenter and νi is Poisson s ratio of the indenter, E* is the reduced modulus. One can find the elastic modulus of the sample, E using Eq. (10). How to Find Hardness, H Hardness, H, is defined by the maximum load divided by the projected area: i 2 i (8) (9) (10) H = P A max (11) 21

22 where Pmax is peak load and A is projected area of contact at the peak load. The unit of hardness is given in N/m 2 =Pa. Materials and Sample Preparation Performance Grade (PG) asphalt binder sample was collected from the local contactor associated with the New Mexico Department of Transportation. PG binder sample was used in this study. Figure 14 shows a laboratory prepared asphalt binder film on glass substrate. As the first step, a glass slide surface 12.7 mm 12.7 mm was selected and weighed in scale up to 4 significant decimal digits of grams. Next, the glass slide was wrapped with high temperature resistant tape. The tape was placed so that it formed the 6.35 mm square gap area previously outlined for the binder. Then, hot polymer modified liquid asphalt binder was poured into the gap of the tape strips. The polymer-modified binders were melted by heating them to 163 o C for an hour. The asphalt coated surface was placed in the oven at 163 o C for 10 min in order to have a smooth surface, cooled at room temperature and the tapes were removed. Finally, the glass slide with the asphalt coating was weighed again to measure the amount of asphalt binder. From the known area, density of the asphalt binder and mass the thickness of the binder film was measured. The film thickness varied within a range of 40 µm to 80 µm to avoid substrate effect on test results. 0.5 inch Glass Substrate 0.5 inch Indentation Location Asphalt Binder Figure 14: Asphalt Binder Sample for Nanoindentation Nanoindentation Test A maximum load of 0.05 mn was applied with an unloading rate of mn/sec. A creep time of 200 sec was applied after reaching the maximum load. A sitting load of 0.01 mn was used for all the samples. The viscous effect of the material like viscoelastic material like asphalt binder can be reduced during nanoindentation by using a fast unloading rate and/or applying an extended dwell time. According to the researchers in viscoelastic material area this approach can eliminate the viscous effect of AC (11-12). It has been shown in the previous study of the authors; if the dwell time is long enough (for asphalt binder) around sec the viscous effect of the material minimizes (23). According to the previous study of the authors a loading rate of 0.02 mn/sec and dwell time of 200 sec was applied for nanoindentation tests of asphalt. However, for the current study pile up effect of indentation has not taken in the account for elastic modulus and hardness 22

23 calculation. It was assumed the fast unloading rate will minimize the effect of pile up around the indenter tip. The nanoindenter device at the UNM nano test laboratory was used for indentation. Figure 15 shows the nanoindentation test setup with the Berkovich indenter tip and sample indenting in asphalt. The pendulum in the system was used to adjust the bridge box output for the Berkovich indenter tip. Figure 15: Nanoindentation Test on Asphalt Binder Results and Discussion In the study, nanoindentation tests were conducted on three different aging condition of PG asphalt binder samples. All the nanoindentation tests were done at 25 ⁰C. Figure 16 shows the load displacement curves obtained from the nanoindentation test on PG binder grade. It shows load displacement curves of for 20 indentations on the binder. The binder showed maximum displacement of 4µm. The figure shows extended dwell time of 200sec produced positively slopped unloading load-displacement curve, which afterwards being analyzed using Oliver-Pharr analysis for determination elastic modulus and hardness of the material. 23

24 Figure 16: Load Displacement Curve Similar load displacement curves were produced from all 24 binder samples and 20 indentations were done to determine the modulus and hardness of the material. Figure 17 shows the modulus of eight different PG binders and 3 different aging conditions. Here it should be noted each column in the figure represents average of 20 indentations. Binder modulus for three different aging condition ranged from 2 to 5 MPa for unaged asphalt binder, 3 to 8 MPa for RTFO aged asphalt binder and 10 to 40 MPa for PAV aged asphalt binders. RTFO aged asphalt binders modulus increases 1.2 to 2.5 times compared to unaged asphalt binder except for PG 64-28, which shows an increase of 6 times from unaged condition. PAV aged asphalt binders shows 7 to 9 times higher modulus compared to unaged asphalt binders modulus except for PG 76-28, which shows an increase of only 2 times compared to unaged binder modulus Unaged RTFO PAV E (MPa) PG PG PG PG PG PG PG PG PG Binder Grade Figure 17: Modulus for different PG binder grades and three different aging conditions 24

25 Binder hardness for three different aging condition ranged from 20 to 60 kpa for unaged asphalt binder, 40 to 170 kpa for RTFO aged asphalt binder and 100 to 440 kpa for PAV aged asphalt binders. Hardness increase for RTFO aged asphalt binder was 1.3 to 4 times compared to unaged asphalt binder, however except hardness of PG for all PAV aging conditions hardness increase 4 to 7 times compared to unaged asphalt binder condition. PG binder shows only 2 times increase in binder hardness compared to unaged asphalt binder modulus of PG In practice, -28 binder grades are usually produced using polymer modification in the asphalt binder industry. The polymer structure might have affected the aged modulus and hardness of and binders. However, from the study the range of indentation drive modulus of all eight binder grades were determined. It is found that irrespective of binder grade the aging of asphalt binder increases the modulus and hardness of the thin film asphalt binder coating. H (kpa) Unaged RTFO PAV PG PG PG PG PG PG PG PG PG Binder Grade Figure 18: Hardness for different PG binder grades and three different aging conditions. Task/Objective 4: Relating G* to E In the previous tasks, The G* and nanoindentation modulus, E of the material was determined for four different asphalt binder grades and three different aging conditions. Though in the current study G* master curve were produced for all 24 asphalt binder samples, to relate both E with G*, the G* of asphalt binder at 22 C at 10 rad/sec was taken. At first the nano scale E is compared with G*of the material. A new parameter n (= E/G*) is used to compare between then. It is found that the nano scale E of the material is 1.5 to 10 times higher that G* of the binder irrespective of the aging condition. The increase in nano scale modulus increase with increase in PG binder grade and aging time. 25

26 12 10 Unaged RTFO PAV 8 n = E/G* PG PG PG PG PG PG PG PG PG Binder Grade Figure 19: Comparison of E and G* However, for PG the effect of aging is not showing any increase in modular ratio parameter. The polymer modification in the binder might have caused reverse effect in modular ratio parameter. Though nano scale E of the binder is always higher than the G* of the material figure 20 shows Pearson relationship of G* and E of the material. G* (MPa) R² = E (MPa) (a) Unaged Condition G* (Mpa) R² = E (MPa) (b) RTFO Aging Condition 26

27 R² = G* (MPa) E (MPa) (c) PAV Aging Condition Figure 20: Pearson Correlation between E and G* of Asphalt Binder Figure 20 shows Pearson correlations of G* and E for three different aging conditions. However, for all three cases R 2 value is less than 0.001, therefore there is no linear correlation exists between nano scale E and Complex modulus G* of asphalt binder. All the indentation drive modulus data and complex modulus data have been introduced in MATLAB to find the nonlinear relationship between E and G* of asphalt binder. The analysis shows 1 st order Gaussian function can reasonably relate nano scale E with micro scale complex modulus G* of asphalt binder. Equation (12) provides the nonlinear Gaussian relationship of E and G*, G E k ( ) m * Ae 2 = (12) where, A, k and m is fitting parameter. A = 4.15 to 6.88 k = to and m = to Figure 21 shows the fitting of equation 12 for all E and G* data obtained in the study. Fitting R squared value for the equation is 0.743, which can be considered as reasonable relationship of E and G* of asphalt binder. 27

28 G* (MPa) Lab Data Model Fit E (MPa) Figure 21: Non Linear Correlation between E and G* of Asphalt Binder. Conclusions The study evaluated the nano scale modulus E and complex modulus G* of eight different asphalt binders commonly used in pavement industry as well as their material property in three different aging condition, i.e. unaged, RTFO aged and PAV aged asphalt binder. Also, a correlation has been established between E and G* of asphalt binder. The following conclusions can be made from this study based on the test results: All the asphalt binders showed increase in G* value with increase in aging in asphalt binder. However, the increase is exponentially higher when binder samples are PAV aged. Both E and H of asphalt binder increase couple of folds compared to unaged asphalt binder sample. RTFO aged asphalt binders modulus increases 1.2 to 2.5 times compared to unaged asphalt binder, whereas PAV aged asphalt binders shows 7 to 9 times higher modulus compared to unaged asphalt binders modulus. Hardness increase for RTFO aged asphalt binder was 1.3 to 4 times compared to unaged asphalt binder, however for all PAV aging conditions hardness increase 4 to 7 times compared to unaged asphalt binder condition. Irrespective of aging condition and binder grade nano scale E of the binder is always 1.5 to 10 times higher than corresponding G* of the binder. There is a nonlinear Gaussian correlation exists between E and G* of asphalt binder. The model showed good relationship agreement between E and G* of asphalt binder. Recommendations Asphalt binder is a viscoelastic material. The current study showed a good nonlinear relationship between E and G* of asphalt binder. However, the nano scale creep response and the creep response in DSR would provide a good relationship between the material 28

29 properties in two different scales. Therefore, research is required to determine the creep response of asphalt binder. 29

30 References 1. Raquel, M., Velasquez, R., and Bahia, H. U. (2011) Measuring the Effect of Moisture on Asphalt-Aggregate Bond with the Bitumen Bond Strength Test. Transportation Research Record, 2209, Washington, D.C., pp Huet, C. (1965) Etude par une methode d'impedance, du compormement viscoelastique des materiaux hydrocarbones (Study of viscoelastic behavior of bituminous mixes by method of impedance). Annales des Ponts et Chaussees, Vol. 135, No.6, pp Little, D. N., Letton, A., Prapnnachari, S. and Kim, Y. R. (1995) Rheological and rheooptical characterization of asphalt cement and evaluation of relaxation properties. Transportation Research Record, 1436, Washington, D.C., pp Masad, E., Somadevan, N., Bahia, H. U. and Kose, S. (2001) Modeling and experimental measurements of strain distribution in asphalt mixes. Journal of Transportation Engineering, ASCE, Vol. 127, No. 6, pp Buttlar, W. G., and You, Z. (2001) Discrete element modeling of asphalt concrete: Microfabric approach. Transportation Research Record, 1757, Washington, D.C., 2001, pp Guddati, M. N., Feng, Z., and Kim, Y. R (2002) Toward a micromechanics-based procedure to characterize fatigue performance of asphalt concrete. Transportation Research Record, 1789, Washington, D.C., 2002, pp Xu, Q., and Solaimania, M. (2009) Modelling linear viscoelastic properties of asphalt concrete by the Huet-Sayegh model. International Journal of Pavement Engineering, Vol. 10, No. 6, 2009, pp Jager, A., Lackner, R. and Sangl K. (2010) Microscale characterization of bitumenback analysis of viscoelastic properties by means of nanoindentation. International Journal of Material Research (formerly Z. Metallkd.), Vol. 98, pp Zofka, A., and Nener-Plante, D. (2011) Determination of Asphalt Binder Creep Compliance Using Depth-Sensing Indentation. Experimental Mechanics, Vol. 51, No. 8, pp Tarefder, R. A., Zaman, A. M. and Uddin, W. (2010) Determining Hardness and Elastic Modulus of Asphalt by Nanoindentation. Journal of Geomechanics, ASCE, Vol. 10, No. 3, pp Tarefder, R. and Faisal, H. (2013) Effects of Dwell Time and Loading Rate on the Nanoindentation Behavior of Asphaltic Materials. Journal of Nanomechanics and Micromechanics, ASCE, Vol. 3, No. 2, pp Tarefder, R. A. and Faisal, H. M. (2013) Nanoindentation Characterization of Asphalt Concrete Aging. Journal of Nanomechanics and Micromechanics, ASCE, 2013, Vol.4, No.1, A Tarefder, R. A. and Faisal, H. M. (2013) Modeling Nanoindentation Creep Behavior of Asphalt Binder. Advances in Civil Engineering Materials, ASTM, Vol. 2, No.1, 14. Ossa, E. A., and Collop, A. C. (2007) Spherical indentation behavior of asphalt mixtures. Journal of Material in Civil Engineering, Vol. 19, No. 9, pp Ossa, E. A., Deshpande, V. S. and Cebon, D. (2005) Spherical indentation behaviour of bitumen. Acta Materialia, Vol. 53, No. 11, pp Woo, W. J., Chowdhury, A. and Glover, C. J. (2008) Field Aging of Unmodified Asphalt Binder 30

31 in Three Texas Long-Term Performance Pavements. Journal of Transportation Research Board No. 2051, Washington, D.C., pp Singh, D., Zaman, M. and Commuri, S. (2011) Effect of Long Term Oven Aging on Dynamic Modulus of Hot Mix Asphalt. Geo-Frontiers 2011: Advances in Geotechnical Engineering, Vol. GSP 211, 2011, pp Huang, S. C. and Turner, T.F. (2014) Characteristics of RAP Blend Binders: Rheological Properties. Journal of Materials in Civil Engineering, ASCE, Vol.26, No. 5, pp Tarefder, R. A., Faisal, H. M and Sobien, H. (2013) Nanomechanical Characterization Effect of Mica and Aging on Asphalt Binder. Journal of Materials in Civil Engineering, ASCE, 2013, /(ASCE)MT , Faisal, H. M., Tarefder, R. A. and Weldegiorgis, M. T. (2014) Characterization of Asphalt Mastic and Aggregate Phases for Moisture Damage. Transportation Research Board, Oliver, W. C., and G. M. Pharr. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. Journal of Material Research, Vol. 7, No. 6, 1992, pp Tarefder, R. and H. Faisal. Effects of Dwell Time and Loading Rate on the Nanoindentation Behavior of Asphaltic Materials. Journal of Nanomechanics and Micromechanics, ASCE, Vol. 3, No. 2, 2013, pp Tarefder, R. A. and H. M. Faisal. Modeling Nanoindentation Creep Behavior of Asphalt Binder. Advances in Civil Engineering Materials, ASTM, 2013, Vol. 2, No. 1, pp Faisal, H. M., M. I. Hossain, and R. A. Tarefder. Nanomechanical Evaluation of Vapor Conditioned and Unconditioned Asphalt. Transportation Research Record: Journal of the Transportation Research Board, No. 2506, Transportation Research Board, Washington, D.C., 2015, pp