Investigations into the relationship between binder ...

2 downloads 0 Views 424KB Size Report
All binders used in this investigation were samples of plant manufactured product. ... original work was then subjected to heat treatment in the laboratory (28 days heating at ..... Tushan Fernando and Andrew Scott of the BP Bitumen Altona.

[Home] [Session index] [Author index] [Presenter index] th

24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

INVESTIGATIONS INTO THE RELATIONSHIP BETWEEN BINDER CONSISTENCY/UNDERLYING VISCOSITY AND ASPHALT RUTTING Robert Urquhart, Mick Budija and Graham Wilson BP Australia Pty. Ltd., Australia ABSTRACT A major failure mode for roads in Australia is rutting of the pavement under heavy traffic loads. Remediation of rutted pavements increases costs for asset owners and impacts the environment as new materials and energy are required for road maintenance. Consistency and Underlying Viscosity at 60°C are binder properties determined using the ARRB Elastometer. Several studies have looked for a relationship between these properties and asphalt wheel tracking which is the laboratory measure for rutting. Each study has used a different asphalt mix design to determine wheel tracking performance. Correlations between Elastometer properties and wheel tracking performance have been observed in some studies but not others, suggesting that the relationships observed are mix design dependent. This paper explores the relationship between Elastometer properties and wheel tracking results for a range of binders in two dense graded mix designs. These results and those reported in the literature have been analysed to establish whether a mix design independent correlation exists between Elastometer properties and wheel tracking performance.

INTRODUCTION A major mode of failure for roads in Australia is permanent deformation or rutting of the pavement under heavy traffic loads. Rutting can occur in one or more of the asphalt layers in the pavement or can extend throughout the entire road and into the subgrade. Rutting can be the result of moving or stationary traffic. It particularly occurs where the road is subject to high shearing stresses that occur as the result of braking, accelerating or turning traffic. A major concern of road asset owners is the financial and environmental costs associated with remediation of rutted pavements. Environmental costs include the use of new materials and energy to repair the rutted pavement. The Australian Road Research Board (ARRB) Group Ltd conducted an extensive study in the late 1990s which investigated the relationship between a number of different binder test properties and asphalt rutting as determined by laboratory wheel tracking tests (Oliver 1997). The investigation involved determining the binder properties and laboratory wheel tracking performance of 10 different binders in a single laboratory prepared Size 14 dense graded asphalt mix which was typical of road authority designs. The main conclusion of this work was that there was a good correlation between the Consistency of the binder (as measured by the ARRB Elastometer) and wheel tracking performance of the final asphalt mix. It was therefore recommended that binder Consistency could be used for specification purposes so that the purchaser could be provided with assurance that supplied polymer modified binders (PMBs) would show satisfactory rutting performance on the road. Following the original ARRB work, a number of industry studies were conducted which investigated the relationship between binder Consistency at 60°C and laboratory wheel tracking test performance (Tredrea 2003). These studies used different asphalt mix designs from the original ARRB work and predominantly focussed on determining whether there was a relationship between binder Consistency at 60°C and wheel tracking performance for highly modified styrene-butadiene-styrene (SBS) based PMBs which were subjected to different amounts of heat treatment. No correlation between binder Consistency at 60°C and wheel © ARRB Group Ltd and Authors 2010

1

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

tracking performance was found in the industry studies. In order to account for these results, Tredrea (2003) defined a new parameter, Underlying Viscosity (50-100%) at 60°C, which was derived by extrapolation of the raw results in the Consistency test in the Strain region between 0.5 and 1.0. A recent report by Choi (2010) on another extensive study into the relationship between binder properties and asphalt rutting included a data set of 12 binders which were subjected to laboratory wheel tracking tests in a Size 10 dense graded asphalt mix. The conclusions of this investigation were that neither binder Consistency at 60°C nor Underlying Viscosity (50-100%) at 60°C appeared to show a good correlation with wheel tracking performance in this mix. Alternatively, Choi found that another parameter derived from binder Consistency measurements, referred to as Underlying Viscosity (20-40%) at 60°C, did appear to show a correlation with wheel tracking performance. However, no correlation was found between these two parameters when Remtulla, Chik and Man (2009) investigated a series of binders in a Size 10 dense graded asphalt mix that met Roads and Traffic Authority of New South Wales requirements. This suggests the relationship between wheel tracking and Underlying Viscosity (20-40%) may not be applicable to all asphalt mix designs. This paper investigates the relationship between binder Consistency at 60°C, Underlying Viscosity (20-40%) at 60°C and laboratory asphalt wheel tracking results for a range of binders in Size 10 and Size 14 asphalt mixes which meet State Road Authority requirements. The results obtained for these two mix designs, as well as those reported in the literature, have been analysed to determine whether a mix design independent correlation exists between Elastometer properties and wheel tracking performance.

MATERIALS AND TEST METHODS Binders All binders used in this investigation were samples of plant manufactured product. Polymer modified binders (PMBs) contained either styrene-butadiene-styrene (SBS), polybutadiene (PBD) or ethylene-vinyl-acetate (EVA) polymers. Binders characterised in the Vicroads Size 10 asphalt mix were used as received. For work using the Queensland Department of Transport and Main Roads (QDTMR) Size 14 asphalt mix, a range of binders (C320 bitumen, A25E, A20E and A15E binders) were initially tested as received. The sample of the A15E binder used in the original work was then subjected to heat treatment in the laboratory (28 days heating at 180°C, followed by 24 days heating at 205°C) so that a range of samples could be tested after different periods of heat treatment. Binders that are representative of grades listed in Australian Standard AS2008 (Standards Australia 1997) or the Austroads AP-T41/06 Specification (Austroads 2006a) are shown in the figures.

Binder tests Consistency at 60°C tests were performed according to the Austroads test method AG:PT/T121 (Austroads 2006b). In order to ensure a direct comparison of the results between binders, all Consistency tests were conducted with Mould A using a test speed of 1.0 mm/s and a breakpoint of 10.0 mm. These parameters correspond to using a Strain Rate of 0.1 s-1 and a final Strain of 1.0 in each test. Each binder was subjected two Consistency tests. The values reported in this paper correspond to the average of the two values determined on each binder. Underlying Viscosity (20-40%) at 60°C results were calculated using the method of Choi (2010). They were derived by initially converting the data from each raw Consistency file to plots of Stress (in Pa) versus Strain. Underlying Viscosity (20-40%) was calculated by determining the intercept at Strain = 0 of a linear fit to the Stress-Strain plot in the Strain region between 0.2 and 0.4 and then dividing the intercept value by the strain rate used in the test (0.1 s-1). As all binders were subjected to two Consistency tests, the values of Underlying Viscosity (20-40%) reported in this paper are the average of the two results obtained for each binder.

© ARRB Group Ltd and Authors 2010

2

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

Asphalt Wheel Tracking Tests Asphalt wheel tracking tests were performed at 60°C according to Austroads test method AG:PT/T231 (Austroads 2006c) using either a mix that conformed to the aggregate grading requirements of a Vicroads Size 10, Type N mix (Vicroads 2008) or a QDTMR DG14 mix (Queensland Government - Department of Main Roads 2006). Table 1 shows the general characteristics of both mixes as well as any deviations from the Standard Reference Test Conditions listed in AG:PT/T231 (Austroads 2006c). Table 1: Characteristics of Vicroads Size 10 and QDTMR Size 14 mixes Characteristic

Vicroads Size 10 mix

QDTMR Size 14 mix

5.4

4.5

4.9 - 6.0

3.8 - 4.7

5.2

4.2

Asphalt binder content (%w/w) Air Void range for all samples tested (%) Average Air Void value over all samples (%)

RESULTS AND DISCUSSION BP Investigations Tables 2 and 3 show values of Consistency at 60°C, Underlying Viscosity (20-40%) at 60°C [UV(20-40%) at 60°C] and Final Wheel Tracking Depth for the range of binders investigated in Vicroads Size 10 and QDTMR Size 14 asphalt mixes. As many of the binders did not show marked differences in Final Wheel Tracking Depth even though they had quite different Consistency and UV(20-40%) results (e.g. wheel tracking results for A20E, A15E, A10E and A35P binders in the Vicroads mix, and heated A15E binders in the QDTMR mix) calculations were performed to determine whether the small differences observed between binders were significant. As no estimate of test precision has been determined for AG:PT/T231 (Austroads 2006c), the error in wheel tracking results was calculated at the 95% confidence limit using a method used in quantitative chemical analysis (Jeffrey 1989) where the error in a test result obtained by averaging separate repeat experiments (in this case 2 or 3 separate wheel tracking tests using each binder) can be calculated if the number of replicate tests and the standard deviation of replicate test results are known. Tables 2 and 3 list the error in each Final Wheel Tracking Depth determined by this method. Table 2: Test results for binders characterised in the Vicroads Size 10 asphalt mix Binder Grade

Consistency at 60°C (Pa s)

Underlying Viscosity (20-40%) at 60°C (Pa s)

Final Wheel Tracking Depth (mm)

95% Confidence Limit of Final Wheel Tracking Depth (mm)

C320 bitumen

544

535

8.0

± 2.5

C600 bitumen

707

696

6.0

± 2.8

C600 bitumen

654

582

7.1

± 0.8

C600 bitumen

650

568

7.1

± 1.8

M1000/320

2052

1411

4.4

± 1.3

A20E

3196

708

3.3

± 0.7

A15E

8027

1081

2.9

± 0.7

A10E

12704

1298

2.5

± 0.5

A35P

2845

1852

2.5

± 0.3

© ARRB Group Ltd and Authors 2010

3

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

Table 3: Test results for binders characterised in the QDTMR Size 14 asphalt mix Binder Grade

Heating time (days)

Consistency at 60°C (Pa s)

Underlying Viscosity (20-40%) at 60°C (Pa s)

Final Wheel Tracking Depth (mm)

95% Confidence Limit of Final Wheel Tracking Depth (mm)

C320 bitumen

0

351

3511

3.32

± 0.7

1

A25E

0

1223

N/A

1.8

± 0.5

A20E

0

3463

1118

1.3

± 0.3

A15E

0

12150

2622

1.1

± 0.2

A15E

4

8764

1875

1.1

± 0.4

A15E

7

7982

1710

1.0

± 0.4

2

± 1.5

A15E

10

7269

1545

1.0

A15E

14

7054

1434

0.8

± 0.6

1484

2

± 0.6

A15E

52

3785

1.2

1

UV(20-40%) at 60°C of the C320 bitumen sample was set to the same value as its Consistency at 60°C. UV(2040%) at 60°C of the A25E sample was not determined as the raw Consistency data files were no longer available.

2

Average of two wheel tracking tests. All other data is the average of three wheel tracking tests.

Choi (2010) reported a linear relationship between Final Wheel Tracking Depth and UV(20-40%) at 60°C for a set of 12 binders in a Size 10 asphalt mix when plotted on a log-log scale. Figure 1 shows the plots obtained when Choi’s analysis is employed using the current data. The relationship between Final Wheel Tracking Depth and UV(20-40%) is adequate for the data obtained for the QDTMR Size 14 mix with a R2 value of 0.82 for the linear fit. However, the value of R2 obtained for the Vicroads Size 10 mix data is low (0.68) indicating a poor relationship between Final Wheel Tracking Depth and UV(20-40%) for this mix. This lack of correlation can also be seen in Table 2 where the first C600 bitumen binder listed in the table and the A20E binder have almost identical UV (20-40%) values (696 and 708 Pa s, respectively) but quite different Final Wheel Tracking Depth (6.0 and 3.3 mm respectively). The M1000/320 and A10E binders also have similar UV (20-40%) results in Table 1 (1411 and 1298 Pa s, respectively) but the Final Wheel Tracking Depth of the M1000/320 binder is more than 1.7 times that of the A10E binder. The results obtained for the Vicroads Size 10 mix indicate that the relationship between Final Wheel Tracking Depth and UV(20-40%) observed by Choi (2010) does not apply to all asphalt mix designs. Studies by Remtulla, Chik and Man (2009) also found no relationship between Final Wheel Tracking Depth and UV(20-40%) when the wheel tracking performance of an A10E binder subjected to different amounts of heat aging was tested in a Roads and Traffic Authority (RTA) of New South Wales Size 10 dense graded asphalt mix. The correlation observed by Choi therefore may be specific to the mix design studied.

© ARRB Group Ltd and Authors 2010

4

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

1.2

log10(Final Wheel Tracking Depth (mm))

y = -0.85x + 3.14 2 R = 0.68

VicRoads Size 10 mix

1.0

QDMR Size 14 mix

C320

0.8 C600 0.6

C320

M1000/320

A20E A15E

0.4 y = -0.63x + 2.07 2 R = 0.82

0.2

A35P A10E

A15E

A20E

0.0 2.4

2.6

2.8

3.0

3.2

3.4

3.6

-0.2 log10(Underlying Viscosity (20-40%) at 60°C (Pa s))

Figure 1: Final Wheel Tracking Depth at 60°C versus Underlying Viscosity (20-40%) at 60°C for binders in Vicroads Size 10 and QDTMR Size 14 mixes. Solid line: Linear fit to loglog data for Vicroads Size 10 mix. Dashed Line: Linear fit to log-log data for QDTMR Size 14 mix. With the less than desirable correlation obtained using Choi’s log-log relationship, other treatments of the data were investigated. Figure 2 shows the result of plotting the wheel tracking data against the inverse of the UV(20-40%) data. This improves the correlations with R2 values of 0.76 for the Vic Roads Size 10 mix and 0.96 for the QDTMR Size 14 mix.

Vicroads Size 10 Mix

9.0

C320

8.0

Final Wheel Tracking Depth (mm)

QDTMR Size 14 Mix

C600

7.0

y = 3771.2x + 0.213 R2 = 0.7558

6.0

5.0

M1000/320 4.0

C320 A20E

3.0

A15E

A35P A10E

2.0

1.0

A15E

y = 968.95x + 0.4519 R2 = 0.9636

A20E

0.0 0

0.0005

0.001

0.0015

0.002

0.0025

0.003

-1

1/Underlying Viscosity (20-40%) ((Pa s) )

Figure 2: Final Wheel Tracking Depth at 60°C versus Inverse Underlying Viscosity (2040%) at 60°C for binders in Vicroads Size 10 and QDTMR Size 14 mixes. Solid line: Linear fit to data for Vicroads Size 10 mix. Dashed Line: Linear fit to data for QDTMR Size 14 mix. However, the best correlations were obtained when the wheel tracking data was plotted against the inverse of the Consistency data. Such a treatment provided R2 values of 0.96 for the Vicroads Size 10 mix and 0.98 for the QDTMR Size 14 mix as shown in Figure 3. © ARRB Group Ltd and Authors 2010

5

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

Vicroads Size 10 Mix

9.0

QDTMR Size 14 Mix C320

Final Wheel Tracking Depth (mm)

8.0

C600

7.0

y = 3037.6x + 2.291 R2 = 0.9587

6.0

M1000/320

5.0

4.0

A20E

C320

A15E 3.0

A10E

2.0

A35P

A25E

y = 832.23x + 0.9411 R2 = 0.9761

1.0

A15E

0.0 0

A20E 0.0005

0.001

0.0015

0.002

0.0025

0.003

-1

1/Consistency ((Pa s) )

Figure 3: Final Wheel Tracking Depth at 60°C versus Inverse Consistency at 60°C for binders in Vicroads Size 10 and QDTMR Size 14 mixes. Solid line: Linear fit to data for Vicroads Size 10 mix. Dashed Line: Linear fit to data for QDTMR Size 14 mix. By viewing the wheel tracking versus Consistency data in a different form reveals other aspects of the relationship. Figures 4 and 5 show plots of Final Wheel Tracking Depth versus Consistency for the range of binders investigated in Vicroads Size 10 and QDTMR Size 14 asphalt mixes. It is evident from these plots that for these two asphalt mixes the Final Wheel Tracking Depth changes rapidly with Consistency up to a critical level. Beyond this point large changes in binder Consistency have minimal effect on wheel tracking performance. This is further highlighted by the A15E ageing study where an A15E binder was kept in heated storage for up to 52 days (see Table 3 and the unfilled data points in Figure 5). Although there is a change in Consistency of over 8000 Pa s, there is no change in the Final Wheel Tracking Depth. In the case of these two asphalt mixes the critical Consistency point seems to be around 3000 Pa s. The Final Wheel Tracking Depth versus Consistency data has been fitted to an equation of the form:

A Final Wheel Tracking Depth (mm) =

Consistency at 60°C (Pa s)

+ B

equation (1)

where A and B are fitted parameters obtained for a specific asphalt mix. The value of B corresponds to the limiting Final Wheel Tracking Depth for the mix at high Consistency at 60°C values. The vertical error bars in the plots correspond to the error in each Final Wheel Tracking Depth result at the 95% confidence limit. If the error in test results is considered then there is not a significant difference in Final Wheel Tracking Depth for A35P, A20E, A15E or A10E binders in Figure 4 or A20E, A15E and all heated A15E binders in Figure 5.

© ARRB Group Ltd and Authors 2010

6

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

12

A = 3038 B = 2.29 R2 = 0.96

Unheated binders Fit to equation (1)

Final Wheel Tracking Depth (mm)

10 C320 8 C600 6

M1000/320 4 A20E 2

A35P

A15E

A10E

0 0

2000

4000

6000

8000

10000

12000

14000

Consistency at 60°C (Pa s)

Figure 4: Final Wheel Tracking Depth at 60°C versus Consistency at 60°C for a range of binders in a Vicroads Size 10 mix. The fitted curve and values of A, B and R2 are a fit of the experimental data to equation (1).

Final Wheel Tracking Depth (mm)

6

Unheated binders Heated A15E binder Fit to equation (1)

5

A = 832 B = 0.94 R2 = 0.98

4 C320

3

2

A25E

1

A20E

A15E

0 0

2000

4000

6000

8000

10000

12000

14000

Consistency at 60°C (Pa s)

Figure 5: Final Wheel Tracking Depth at 60°C versus Consistency at 60°C for a range of unheated and heated binders in a QDTMR Size 14 mix. The fitted curve and values of A, B and R2 are a fit of the experimental data to equation (1)

Investigations by other workers As the results of this and other investigations did not show a relationship between Final Wheel Tracking Depth and UV(20-40%) at 60°C, but equation (1) showed excellent fits to both sets of BP results, data from a number of similar investigations that have been reported in the literature were analysed to determine whether equation (1) could be used to describe the results obtained in other mix designs. In order to properly assess whether equation (1) could be used to fit the data, only studies which covered a number of different binders and had published details of the © ARRB Group Ltd and Authors 2010

7

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

asphalt mix design used are included in the data below. Studies in which asphalt compaction was not done in the laboratory were also excluded (e.g. Oliver et al. 1995) to ensure that air void variations (which can occur if asphalt slabs containing different binders are taken directly from a finished pavement) were not contributing to any correlation, or lack of correlation, between the experimental data and equation (1). Figure 6 shows the results of Oliver (1997) that were obtained from a range of binders in a Size 14 mix which was designed to show enhanced asphalt rutting (A. Alderson, pers comm.). The Steady State Wheel Tracking Rates shown in the figure have been converted from mm/kcycle to mm/kpass so that they align with the values reported in Austroads test method AG:PT/T231 (Austroads 2006c). During Oliver’s study, the Consistency of ten different binders was measured at 45°C and 60°C as well as their wheel tracking performance at both temperatures. This experimental method allowed Oliver to obtain a data set of 20 different samples with a wide range of Consistency values when wheel tracking data was matched with Consistency data at the same temperature.

Steady State Wheel Tracking Rate at 60°C or 45°C (mm/kpass)

12 C170

Oliver data A = 1804 B = 0.087 R2 = 0.98

Oliver data conducted at 60°C

10

Oliver data conducted at 45°C Oliver data: Fit to equation (1)

8

6

C320

4 C600 M1000/320 C170 C320

2

M1000/320

C600

0 0

5000

10000

15000

20000

25000

30000

35000

40000

Consistency at 60°C or 45°C (Pa s)

Figure 6: Steady State Wheel Tracking Rate versus Consistency for a range of binders in a Size 14 dense graded asphalt mix that was typical of road authority designs. The binder content of the mixes was 5.0% w/w. The fitted curve and values of A, B and R2 are a fit of the experimental data to equation (1). The data in Figure 6, in an analogous way to that of Figures 1 and 2, shows that there is an abrupt reduction in Steady State Wheel Tracking Rate with binder Consistency for Consistency values less than about 4000 Pa s. For Consistency values greater than 4000 Pa s, there is essentially no change in Steady State Wheel Tracking rate with binder Consistency. Several workers (Choi 2010; van Loon, Figueroa & Simpson 2008) have found that for a single asphalt mix there is a strong correlation between Steady State Wheel Tracking Rate and Final Wheel Tracking Depth. In agreement with their observation, equation (1) was equally successful at fitting Oliver’s Steady State Wheel Tracking Rate data as it had been for BP’s Final Wheel Tracking Depth data in the previous section. Figure 6 shows the values of A, B and R2 obtained by fitting the Steady State Wheel Tracking Rate data. As can be seen from the figure, equation (1) gives an excellent fit to Oliver’s data. Figure 7 shows the data from Choi (2010) that were obtained in a Size 10 dense graded asphalt mix. There is again a marked reduction in Final Wheel Tracking Depth with Consistency at 60°C for Consistency at 60°C values less than about 2000 Pa s. With the exception of the binder labelled “PMB F”, further increases in binder Consistency at 60°C do not have a marked effect on wheel tracking performance. Choi (2010) needed to modify the wheel tracking test in order to obtain a result for PMB F as the slab made with this material was so sticky that a large © ARRB Group Ltd and Authors 2010

8

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

amount of material stuck to the test wheel when standard wheel tracking test conditions were used. Choi’s modification to the test, which involved placing a thin rubber membrane over the asphalt slab during testing, may be the reason why the result obtained for PMB F differs from other results obtained in his study.

14

Choi data Final Wheel Tracking Depth (mm)

12

Choi data A = 2195 B = 1.98 R2 = 0.85

Fit to equation (1) C170

10 C320 S35E

8 6 4

PMB F 2

C600

M1000/320

A15E

A35P

A10E

0 0

2000

4000

6000

8000

10000

12000

14000

Consistency at 60°C (Pa s)

Figure 7: Final Wheel Tracking Depth at 60°C versus Consistency at 60°C for 12 different binders in a Size 10 dense graded asphalt mix. The binder content of the mixes was 6.8% w/w. The fitted curve and values of A, B and R2 are a fit of the experimental data to equation (1). Figure 7 shows the values of A, B and R2 obtained by fitting the experimental data to equation (1). The value of R2 obtained from the plot (0.85) is indicative of a good fit to the experimental data. (Excluding the PMB F data point improves the fit to an R2 of 0.88.) This value of R2 is similar to that obtained by Choi (0.88) when he fitted his data and found a relationship between Final Wheel Tracking Depth and UV(20/40) at 60°C. Figure 8 shows plots of Final Wheel Tracking Depth versus Consistency at 60°C for two different mix designs that were reported by Remtulla, Chik and Man (2009) and Maccarrone, Ky and Gnanaseelan (1996). As no Consistency at 60°C data was available for the C170 and C320 binders in these studies, values for these materials were set to 170 and 320 Pa s, respectively. The Final Wheel Tracking Depth result for the C170 binder was obtained from Tredrea (2002). Maccarrone’s study involved testing of unheated binders. The study by Remtulla involved testing an unheated C170 binder and samples from a 20 tonne batch of A10E binder which was manufactured and stored in a tank at 185°C for up to 34 days. Samples were taken from the batch just after manufacture as well as after various times of heating. The result obtained from the sample taken just after manufacture is labelled as the A10E binder in the figure.

© ARRB Group Ltd and Authors 2010

9

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

14

Remtulla data Final Wheel Tracking Depth (mm)

12

Remtulla data A = 228 B = 1.95 R2 = 0.94

Remtulla data: Fit to equation (1)

C320

Maccarrone data Maccarrone data: Fit to equation (1)

10

Maccarrone data A = 3077 B = 2.39 R2 = 0.99

8 6 M1000/320 4 2 C170

A10E

0 0

5000

10000

15000

20000

25000

30000

35000

Consistency at 60°C (Pa s)

Figure 8: Final Wheel Tracking Depth at 60°C versus Consistency at 60°C for two Size 10 dense graded asphalt mixes. Remtulla mix: Size 10 mix meeting the Roads and Traffic Authority of New South Wales R116 Specification. Binder Content = 5.8% w/w. Maccarrone mix: Size 10 mix compacted to 5% air voids. The fitted curves and values of A, B and R2 are fits of the experimental data to equation (1). The data of Remtulla, Chik and Man (2009) shows the same trends observed for other asphalt mix designs described in this paper as equation (1) fits the experimental data extremely well and the Final Wheel Tracking Depth does not change with binder Consistency at 60°C for Consistency at 60°C values greater than 3000 Pa s. Even though the data of Maccarrone, Ky & Gnanaseelan (1996) only consists of three data points, equation (1) describes the data extremely well and the results show a marked reduction of Final Wheel Tracking Depth with Consistency at 60°C for low binder Consistency at 60°C values. For all asphalt mix designs shown in this paper there appears to be a marked reduction in Final Wheel Tracking Depth or Steady State Wheel Tracking Rate with binder Consistency at 60°C for Consistency values less than 2000 - 4000 Pa s. If the Consistency at 60°C of the binder is above these values there is no marked change in the wheel tracking performance of the binder. Consistency at 60°C is included in the Austroads AP-T41/06 Specification for PMBs (Austroads 2006a) as it is meant to provide assurance to the purchaser that supplied products will not rut significantly when they are placed on the road. As all the asphalt mix designs described in the paper are typical of that used in road construction in Australia, a Consistency at 60°C specification of at most 4000 Pa s minimum for even the most highly modified PMB binders should be sufficient to ensure that an asphalt pavement layer could be considered to be resistant to rutting.

CONCLUSIONS The results obtained in this investigation in conjunction with those obtained by Remtulla, Chik and Man (2009) clearly show that the correlation between Final Wheel Tracking Depth and Underlying Viscosity (20-40%) at 60°C observed by Choi (2010) does not apply to all asphalt mix designs. A relationship between Final Wheel Tracking Depth or Steady State Wheel Tracking Rate and Consistency was found which provided excellent fits to the data obtained using a wide variety of binders in 6 different asphalt mix designs. In all systems described, there was a marked drop in Final Wheel Tracking Depth or Steady State Wheel Tracking Rate with Consistency at 60°C for binders with low Consistency at 60°C values. Further increases in wheel tracking performance were not observed if the Consistency at 60°C of the binder was above the range of 2000-4000 Pa s. These results suggest that even for the most highly

© ARRB Group Ltd and Authors 2010

10

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

modified PMB binders, a binder Consistency at 60°C result above 4000 Pa s is sufficient to ensure that the final asphalt layer will be resistant to rutting.

ACKNOWLEDGEMENT We thank John Papadopoulos, Tushan Fernando and Andrew Scott of the BP Bitumen Altona Technical Centre for conducting the Consistency at 60°C tests on the samples described in this paper.

REFERENCES Austroads 2006a, AP-T41/06, Specification Framework for Polymer Modified Binders and Multigrade Bitumens, Sydney, NSW. Austroads 2006b, AG:PT/T121, Consistency, Stiffness, Elastic Recovery and Tensile Modulus of Polymer Modified Binders (ARRB Elastometer), Austroads, Sydney, NSW. Austroads 2006c, AG:PT/T231, Deformation Resistance of Asphalt Mixtures by the Wheel Tracking Test, Austroads, Sydney, NSW. Choi, Y 2010 , Laboratory Study on Relationship between Binder Properties and Asphalt Rutting, Austroads Project No. TT1354. Jeffrey, GH 1989, Vogel’s textbook of quantitative chemical analysis, 5th Edition, Longman Scientific and Technical, United Kingdom, p 138. Maccarrone, S, Ky, A & Gnanaseelan, GP 1996, ‘Rutting and Fatigue Properties of High Performance Asphalt’, Proceedings Roads 96 Conference, Part 2. Christchurch, New Zealand, pp. 133-147 Oliver, JWH, Alderson, AJ, Tredrea, PF & Karim, MR 1995, Results of the laboratory program associated with the ALF asphalt deformation trial, APRG Report No. 12, research report ARR 272, ARRB Transport Research, Vermont South, Vic. Oliver, JWH 1997, ‘Development of PMB Specifications for Asphalt Rut Resistance’, 10th APPA International Flexible Pavements Conference, Perth, Australia, Australian Asphalt Pavement Association (AAPA), Hawthorn, Victoria, 20 pp. Queensland Government - Department of Main Roads 2006, MRS11.30 Main Roads Standard Specification - Dense Graded Asphalt Pavements. Remtulla, A, Chik, B & Man, I 2009, ‘The Effect of Prolonged Hot Storage of the Elastomeric PMBs on the Rutting and Fatigue Properties of Hot Asphalt Mix’, 13th APPA International Flexible Pavements Conference, Surfers Paradise, Australia, Australian Asphalt Pavement Association (AAPA), Hawthorn, Victoria, 10 pp. Standards Australia 1997, AS2008, Residual Bitumen for Pavements. Tredrea, PF 2002, Austroads Project Report, Document RC2013, Progress report on binders research program, ARRB Transport Research, Vermont South, Vic. Tredrea, PF 2003, ‘Viscosity in High Performance Binders and its Contribution to Asphalt Deformation Control’, 21st ARRB and 11th REAAA Conference, Cairns, Australia, ARRB Transport Research Ltd, Australia, 18 pp. van Loon, H, Figueroa, J & Simpson, C 2008, ‘Proposed Wheel Tracking Requirements for Asphalt’, 23rd ARRB Conference - Research Partnering with Practitioners, Adelaide, Australia, ARRB Group, Vermont South, Vic, 16 pp. Vicroads 2008, RC 500.01, Code of Practice for Registration of Bituminous Mix Designs.

© ARRB Group Ltd and Authors 2010

11

th 24 ARRB Conference – Building on 50 years of road and transport research, Melbourne, Australia 2010

AUTHOR BIOGRAPHIES Robert Urquhart completed a PhD in colloid and surface chemistry at the University of Melbourne in 1991. He then held postdoctoral positions at CSIRO (Division of Chemicals and Polymers) and at the University of Minnesota. In 1997 Robert joined BP in the BP Bitumen Technical Centre in Melbourne which conducts product development work and provides manufacturing support to the BP bitumen business in Australia. His first 5 years with BP involved R&D work on bitumen emulsions. From 2002 he managed many R&D projects on bitumen, multigrade and polymer modified bitumens. In 2006 he was appointed to the Technology Services Manager role where he managed the R&D and Quality Control aspects of the BP Bitumen Technical Centre. Mick Budija has 27 years experience in the asphalt and road construction industry specialising in the area of hot mix asphalt, including asphalt mix design, asphalt production and mobile road plant application. Mick worked for Vicroads for 6 years and Pioneer Asphalts (Victoria) for 3 years prior to joining BP Bitumen in 1992. As the Applications Technologist in the BP Bitumen Technical Centre Mick has been involved in many aspects of asphalt technology including asphalt mix design, slurry mix design, and customer and field support. Currently Mick is responsible for all asphalt development work at BP Bitumen. Graham Wilson completed a PhD in Chemistry in 1988 at La Trobe University. He then held a Post Doctoral position in the Physics department at Monash University. Graham moved to South Pacific Tyres in 1990 where he worked on real time data acquisition systems. In 1994, he joined the Kodak Australia Research Laboratory where he led a number of technical teams working on photographic paper systems. Graham joined BP Bitumen as the Technical Centre Manager in 2006. In 2007 he was appointed to the position of Technical Projects Manager and then to Technical Manager in 2010.

© ARRB Group Ltd and Authors 2010

12

Suggest Documents