Methodology for Determining Compaction Temperatures for Modified Asphalt Binders

J. Results for the Mixture With Diabase Aggregate and 1.25-Percent Hydrated Lime

1. Effect of Gyratory Compactor

The data using the diabase aggregate with no hydrated lime indicated that the gyratory compaction process was not always precise. The difference in air voids between the two replicate specimens was sometimes greater than 0.7 percent, even though the effective asphalt binder content did not vary significantly from specimen to specimen or from temperature to temperature.

Three factors may increase the variability in air voids: (a) the amount of age hardening due to the STOA procedure is not always consistent, (b) the process of placing the loose mixture into the gyratory compactor mold is not always consistent, and (c) the gyratory compactor does not consistently apply the same angle. A combination of these factors is also possible. The first reason alone is probably not the answer because the measured allowable compaction temperature ranges were relatively wide. Temperature affects the stiffness of the asphalt binder as does aging.

When the mixture with the PG 58-28 unmodified asphalt binder, diabase aggregate, and 1.25-percent hydrated lime was compacted, the air voids of the replicate specimens varied significantly. The mixture was then compacted using a second gyratory compactor to determine if it could provide more repeatable data. The data are given in table 15. The original gyratory compactor is #1. The data show that neither compactor provided more consistent air voids; therefore, it was decided not to change the compactor. Based on the variability shown in table 15, the number of replicate specimens was increased from two to four. The dry mixing time was also increased. The data for the second pair of specimens using the original compactor are given under "Gyratory Compactor #1 (Repeat)." The new data were more consistent. Increasing the dry mixing time may have allowed more uniform dispersion of the hydrated lime.

Table 10. Effect of compaction temperature on the mixture with
PG 58-28, diabase aggregate, and no hydrated lime.

^aAssumes that the additional asphalt binder will not change the amount of asphalt binder absorption or workability during compaction.

Table 11. Air voids for the mixtures with diabase aggregate and no hydrated lime.

^aThe difference in air voids is large. A difference of 0.7-percent air voids means that the 67- and 95-percent confidence limits for the average asphalt binder content by mass at a 4-percent air-void level will be ±0.25 and ±0.5 percent, respectively, using two replicate specimens.

Table 12. Effect of compaction temperature on the mixture with
Novophalt (PG 76-22), diabase aggregate, and no hydrated lime.

^aAssumes that the additional asphalt binder will not change the amount of asphalt binder absorption or workability
during compaction.

Table 13. Effect of compaction temperature on the mixture with
Styrelf (PG 82-22), diabase aggregate, and no hydrated lime.

^aAssumes that the additional asphalt binder will not change the amount of asphalt binder absorption or workability during compaction.

Table 14. Aggregate gradations before compaction,
after compaction, and after recompaction.

Table 15. Effect of using two different gyratory compactors on the mixture
with PG 58-28, diabase aggregate, and 1.25-percent hydrated lime.

2. PG 58-28 Unmodified Asphalt Binder

Table 16 shows the data for the PG 58-28 unmodified asphalt binder. When the compaction temperature was 157°C, a medium amount of smoke was produced during STOA. Therefore, this temperature cannot be used. The data in the top half of the table were from the four replicate specimens. The data at temperatures of 117, 137, and 157°C each had one outlier that was then removed. The data from the remaining specimens are in the bottom half of table 16. Recommended compaction temperatures are from 107 to 147°C, although the minimum temperature may be lower than 107°C. The air voids for the four replicate specimens are given in table 17.

3. Novophalt (PG 76-22) Polymer-Modified Asphalt Binder

Table 18 shows the data for Novophalt (PG 76-22) after one outlier was removed. When the compaction temperature was 179°C, a large amount of smoke was produced during STOA. Therefore, this compaction temperature cannot be used. Recommended compaction temperatures are from 139 to 159°C. The air voids for the four replicate specimens are given in table 17.

4. Styrelf (PG 82-22) Polymer-Modified Asphalt Binder

Table 19 shows the data for Styrelf (PG 82-22). Recommended compaction temperatures are from 147 to 167°C. The maximum temperature was again based on the amount of smoke produced. The air voids for the four replicate specimens are given in table 17.

Changing the composition of the mastic by adding hydrated lime did not change the ranking for the asphalt binders based on the average compaction temperature. The average compaction temperature for Styrelf was higher than for Novophalt with and without hydrated lime. An interaction between the modified asphalt binders and the aggregate was not evident. Based on this finding, testing mastics with hydrated lime was not necessary and was not done.

K. Results for the Mixtures With Limestone Aggregate

1. PG 58-28 Unmodified Asphalt Binder

Table 20 shows the data for the PG 58-28 unmodified asphalt binder. Recommended compaction temperatures are from 107 to 147°C, although the minimum temperature may be lower than 107°C. The air voids for the four replicate specimens are given in table 21.

2. Novophalt (PG 76-22) Polymer-Modified Asphalt Binder

Table 22 shows the data for Novophalt (PG 76-22). Recommended compaction temperatures are from 119 to 159°C. The air voids for the four replicate specimens are given in table 21.

3. Styrelf (PG 82-22) Polymer-Modified Asphalt Binder

Table 23 shows the data for Styrelf (PG 82-22). Recommended compaction temperatures are from 117 to 167°C, although the minimum temperature may be lower than 117°C. The air voids for the four replicate specimens are given in table 21.

L. Discussion on Recommended Compaction Temperatures

Based on the recommended compaction temperatures given in table 24, a compaction temperature of 145°C could be used for all mixtures. Even so, a procedure for obtaining the compaction temperature range is needed because the ranges for Novophalt and Styrelf will not overlap the range for a PG 46 or PG 52 asphalt binder.

Based on the data from the two diabase mixtures with Novophalt, the inclusion of hydrated lime increased the minimum temperature from 120 to 140°C. A firm conclusion regarding this increase could not be made because the compaction results for both mixtures were highly variable at 140°C. The inclusion of hydrated lime decreased the minimum temperature from 125 to 105°C for the diabase mixtures with the PG 58-28 unmodified asphalt binder. However, if the maximum allowable air-void level for the compaction tests were to be increased from 4.5 to 4.6 percent, both mixtures would provide a minimum temperature of 105°C.

All compaction temperatures for the diabase mixtures with Novophalt and Styrelf, but without hydrated lime, provided air-void levels slightly greater than 4.0 percent. This means that the asphalt binder content would have to be increased to obtain a 4.0-percent air-void level. Four-percent air-void levels were obtained for the diabase mixtures with hydrated lime. This is an example of the complexities that polymer-modified asphalt binders can provide.

The average compaction temperatures given in table 25 show that the temperatures from the asphalt binders and the mastics were high for Novophalt and Styrelf compared to the average temperatures from the Superpave gyratory compactor. Table 26 gives the allowable temperature ranges. The temperature ranges from the viscometers were based on viscosities of 250 and 310 mm²/s, which came from the specified viscosity of 280 ±30 mm²/s. This viscosity range provided a compaction temperature range of 5.0°C. The gyratory compactor provided a significantly higher range of 20°C for all aggregate blends and asphalt binders.

Table 26 shows that the allowable temperature ranges from the capillary viscometers for Novophalt and Styrelf were closer to the ranges from the gyratory compactor than the ranges from the Brookfield viscometer. The viscosities from the Brookfield viscometer should be more accurate because the tests were performed over a range of temperatures that were closer to the temperatures from the gyratory compactor. The marginally better results from the capillary viscometers are most likely the result of errors in the viscosities measured at 60°C. It is also possible that the linear relationship used to obtain the compaction temperatures from the capillary viscosities at 60 and 135°C may not be valid for these modified asphalt binders. Furthermore, the compaction temperatures determined using the capillary viscometers were taken from the relationship between log-log viscosity and temperature, which provided lower temperatures than the relationship using log temperature.

Table 27 shows the viscosities of the asphalt binders at the minimum and maximum allowable compaction temperatures from the Superpave gyratory compactor. Even though a compaction temperature of 145°C could be used for all mixtures, the viscosities of the unmodified and modified asphalt binders at 145°C are not the same.

The Brookfield viscosity range applicable to all aggregate blends is included in table 27. The data from this viscometer were evaluated because it is used by Superpave. The viscosity ranges for the three asphalt binders do not overlap. Therefore, a single viscosity range cannot be specified. However, if the maximum allowable air-void level were to be increased from 4.5 to 4.6 percent, the minimum allowable temperature would be 105°C for all three PG 58-28 mixtures, and the allowable viscosity range would be 200 to 950 mm²/s. The viscosity range applicable to both Novophalt and Styrelf is 900 to 1250 mm²/s. Therefore, a range of 900 to 950 mm²/s is applicable to all mixtures if the maximum air-void criterion is relaxed to 4.6 percent. It is doubtful that this range would be applicable to PG 46 or PG 52 asphalt binder.

M. NCHRP Project 09-10

NCHRP Project 09-10 recently recommended mixing and compaction temperatures for modified asphalt binders.⁽⁹⁾ When the Brookfield viscosities of asphalt binders are measured using the current standardized spindle rate of 20 rpm as in this FHWA study, NCHRP Project 09-10 recommends a target viscosity range of 1300 to 1500 mm²/s. This viscosity range is slightly higher than the range of 900 to 1250 mm²/s provided by the two polymer-modified asphalt binders tested in this study. The compaction temperature range provided by a viscosity range of 1300 to 1500 mm²/s would be 155 to 157°C for Novophalt and 141 to 144°C for Styrelf. The compaction temperature range provided by a viscosity range of 900 to 1250 mm²/s would be 145 to 160°C. Most likely, the latter temperature range is larger because only two polymer-modified asphalt binders were tested.

The compaction temperature range using a viscosity range of 1300 to 1500 mm²/s would be 95 to 100°C for the PG 58-28 unmodified asphalt binder. This range is too low. The current standardized procedure should be used to determine the allowable compaction temperature range for this asphalt binder.

N. Conclusions

• The data strongly suggest that an allowable laboratory compaction temperature range cannot be based on a single asphalt binder viscosity range determined using current standardized procedures.
  
  
• The allowable laboratory compaction temperature range for both polymer-modified asphalt binders tested in this study should be determined using a viscosity of 1100 ±200 mm²/s.
  
  
• The maximum allowable compaction temperature for a given mixture was found to be the highest temperature where the mixture did not produce smoke during STOA. The air voids at this temperature were always in the acceptable range of 4.0 ±0.5 percent. The maximum allowable compaction temperature was then decreased slightly so that smoke would not occur during mixing, which is performed at a higher temperature. The presence of smoke was determined visually; no measurements were taken.

O. Recommendations

• Use the current viscosity of 280 ±30 mm²/s to determine the allowable compaction range for unmodified PG 46, PG 52, and PG 58 asphalt binders.
  
  
• For polymer-modified asphalt binders with a PG of 64 and greater, use a viscosity level of 1100 ±200 mm²/s determined in this study, or a level of 1400 ±100 mm²/s from NCHRP Project 09-10. No recommendation can be given for modified asphalt binders with a PG of 58 or lower. These binders need to be investigated.
  
  
• Investigate the primary procedure recommended by NCHRP Project 09-10 for determining the compaction temperature.⁽⁹⁾ When the Brookfield viscosities of asphalt binders are measured using the current standardized spindle rate of 20 rpm, NCHRP Project 09-10 recommends a target viscosity range of 1300 to 1500 mm²/s. A spindle rate of 20 rpm gives a shear rate of 6.8 s^-1.⁽⁹⁾ However, this is an alternative method. The primary method is to use the temperature at a viscosity of 6000 mm²/s (actually 6.0 Pa-s) determined using a shear rate of 0.001 s^-1. The Brookfield rheometer is not capable of applying this shear rate. Therefore, the viscosity of an asphalt binder at each temperature must be determined by performing the test at various shear rates. A curve-fitting model is then used to calculate the viscosity at 0.001 s^-1.

P. Final Comment

The viscosity range of 1100 ±200 mm²/s (or 1400 ±100 mm²/s) may not be valid when producing specimens with different asphalt binders for mechanical testing comparisons. The compaction temperature range for each asphalt-aggregate combination used in this study was based on obtaining an air-void level of 4.0 ±0.5 percent. No mechanical properties were measured. The compaction temperature ranges for the mixtures may be smaller with less overlap if they were to be based on obtaining the same mechanical properties. Mechanical properties may be more sensitive than density to the differences in age-hardening provided by various compaction temperatures during STOA. The temperature used during STOA is important when comparing mixtures using performance tests.

Previous | Table of Contents | Next