Investigation of Low and High Temperature Properties of Plant Produced Rap Mixtures
OVERALL OBSERVATIONS AND CONCLUSIONS
Based on the results presented in this report, the following observations and conclusions were made:
Test results on binders extracted and recovered from the plant-produced mixes showed that, in general, as the RAP content in the mixture increased, the high-temperature grade of the recovered binder also increased, but only by a few degrees (33.8 to 37.4 1.8 to 5.4 °F (1 to 3 °C)). AMENDED 1/11/2012
As the RAP content increased, the low-temperature grade of the recovered RAP binders also increased, but not as much as the high-temperature grade.
The use of a softer virgin binder grade typically decreased both the high- and low-temperature grades of the recovered binders by half a grade or more.
Increasing the RAP content to 25 percent changed the recovered binder low-temperature grade by no more than 35.6 3.6 °F (2 °C) compared to the binder recovered from the virgin mix (with no RAP). AMENDED 1/11/2012
The low-temperature grades determined by analysis of the BBR data and by the TSAR™ analysis were in close agreement, using the direct tension data and the BBR data to calculate Tcrit.
Examination of the mixture |E*| data reveals that, in general, an increase in the RAP content caused an increase in the modulus of the mix, especially at intermediate and high temperatures. However, this finding was not consistent across every set of mixes, possibly reflecting the important contributions of other factors to the mix modulus.
Statistical analysis of the dynamic moduli at 25 Hz showed that in most cases, there was no significant difference in the moduli of mixes with PG64-22 and varying RAP contents. In cases where there was a statistically significant difference, the 40 percent RAP mix differed from the other RAP contents. These findings suggest that the RAP content can possibly be increased to 25 percent before changing the virgin binder grade, but a grade change should be incorporated when increasing the RAP content to 40 percent. The findings are considered representative of mixtures produced in hot mix plants in Indiana using virgin asphalt binders and RAP typically found in Indiana.
Use of a softer virgin binder grade typically reduced the stiffness of the mixes.
The moduli of the PG58-28 mixes with 25 and 40 percent RAP were often statistically different.
Comparison of the measured mix moduli to those predicted based on the recovered binder master curves and the mixture volumetrics suggests that significant blending of the RAP and virgin binders occurred in 16 of the 20 mixes that contained RAP. Data from contractor 4 showed incomplete blending between the PG64-22 virgin binder and the higher RAP contents. More research is needed to validate this approach to estimating blending.
IDT test results on the mixtures showed only slight effects on Tcrit for mixes containing up to 25 percent RAP and PG64-22. The 40 percent RAP mixes with PG64-22 had warmer (less negative) Tcrit in two cases, and the data did not converge in one case, so the cracking temperature could not be determined. In these mixes, the predicted cracking temperatures were approximately the same as the design low-temperature binder grade (i.e., around -7.6 °F (-22°C)), so would presumably be acceptable.
Tcrit of the mixes with PG58-28 were much lower than those of the comparable mixes with PG64-22 in two of five cases. In the other cases, they were comparable. It is not clear whether cracking temperatures of -14.8 °F (-26 °C) or lower are necessary.
The extraction/recovery process and solvent did not show a clear pattern. That is, binders recovered using the AASHTO T 319 procedure with nPB were not consistently stiffer or softer than those recovered using the Abson procedure with mCl or nPB. As a result, the extraction/recovery/solvent variables do not explain the unexpected results of phase I. Any extraction/recovery procedure is problematic.
The fatigue testing results do not conform to expectations. While the softer virgin binder grade resulted in an increase in the fatigue life at 25 percent RAP, it did not have as strong an impact on the 40 percent RAP mixes.
The mixes with 40 percent RAP exhibited the greatest fatigue lives in many cases, which was an unexpected result.
The fatigue life increased as temperature increased at a given strain level, but fatigue life was less sensitive at lower temperatures.
At the two lower temperatures tested, the predicted fatigue life was the same for all six mixtures regardless of RAP content or virgin binder grade.
More work is needed to investigate the effects of RAP content on fatigue life.
Additional research should be conducted on plant-produced mixtures from different places with varying RAP and virgin material properties to examine the effects of these variables over a wider inference range.
Other agencies should be advised to review their typical materials, especially RAP stockpiles, to determine their own course of action.
Page Owner: Office of Research, Development, and Technology, Office of Infrastructure, RDT