Long-Term Aging of Asphalt Mixtures for Performance Testing and Prediction: Phase III Results (2021)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

120 The main products of the research reported herein are: • New, simplified, and recalibrated CAI that determines laboratory long-term aging durations for loose mix aged in the oven at 95°C to simulate field long-term aging state for the desired climatic conditions and pavement depth; • Refined and field-calibrated PAM that predicts the evolution of binder |G*| at 64°C, 10 rad/s for any field aging duration, climatic conditions, and pavement depth; • A methodology that allows estimating the inputs to the PAM using RTFO and 40 hours of PAV or PG instead of loose mixture aging; and • AMAC model that translates the changes in binder |G*| predicted from the PAM into changes in mixture linear viscoelastic and fatigue properties. Below are some conclusions derived from the work done to reach the products listed above. Refinement of the Climate-Based, Predefined Aging Durations The CAI proposed by Kim et al. (2018) was simplified and recalibrated in this project using both conventional and modern mixtures (e.g., RAP, WMA, and PMA) to yield the required laboratory aging durations as a function of hourly pavement temperature history and depth. The recalibration process aimed to yield laboratory aging durations that offer reasonable accuracy to match field aging levels. The conclusions are as follows: • Outlier behavior observed from many of the PMA sections could have been caused by the unrealistically harsh thermal history used in the ALF rutting experiments. • Roughly half of the PMA sections still fell within the 90% confidence interval of the recali- brated CAI, suggesting promise for the CAI’s applicability to PMA materials. • The recalibrated CAI results represent WMA materials well, suggesting no bias in the required CAI for WMA materials compared to HMA. • The RAP mixtures showed evidence of some difference from the virgin mixtures. How- ever, relatively few RAP mixtures were analyzed, and the corresponding field cores were just 4 years old. • In some instances, the laboratory short-term aged material prepared according to AASHTO R 30 exceeded the age level of the field cores, suggesting that refinement of the short-term aging procedure may be necessary prior to investigating the effect of RAP on the relationship between laboratory long-term aging durations and field aging. C H A P T E R 7 Conclusions and Suggested Research

Conclusions and Suggested Research 121   Refinement of the Pavement Aging Model The PAM was established by calibrating the kinetics model, proposed by Kim et al. (2018), against field core measurements as a function of depth for a wide range of pavement sections, including both conventional HMA and modern materials (i.e., RAP, WMA, and PMA). The conclusions are as follows: • The analysis of the effects of asphalt mixture morphology on aging provided insufficient insights to develop a rigorous diffusion model. • Instead of a diffusion model, depth-dependent field calibration of the kinetics model was carried out using all test sections with asphalt mixtures prepared at the optimum asphalt content. • Sections used in the systematic study of mixture morphology were analyzed to calibrate an adjustment to the PAM predictions as a function of the deviation from the Superpave optimum asphalt content. • An adjustment for air void content was not established because air void content was found to have an insignificant effect on field aging. • The PMA results generally align with the rest of the data, suggesting that a separate field cali- bration of the PAM for PMA materials may be unnecessary. • A limited set of RAP and WMA sections were obtained (up to 4 years of aging). Useful pave- ment sections require the availability of both field cores of long-term aged pavements as well as the component materials used to construct those pavement sections. • Some of the RAP and WMA data were excluded from the field calibration of the PAM. The outlying behavior is attributed to the laboratory short-term aging procedure that overesti- mates field aging in cold climates. • Although not used in the calibration, these RAP sections and WMA sections seemed to follow the same trends as the conventional sections, which indicates that a separate field calibration of the kinetics model for RAP and WMA sections was not required. • For most of the sections evaluated, the PAM predictions agree reasonably well with the field core measurements, except for some of the field sections with RAP. • The PAM predictions generally were found to outmatch the GAS model predictions, with some exceptions. Development of Procedures to Estimate the PAM Inputs Using Standard Binder Aging Methods and Performance Grades The PAM requires two primary material inputs: (1) log |G*| at short-term aging and M and (2) the pavement temperature history. The pavement temperature history can be obtained from the EICM. The material inputs, log |G*| at short-term aging and M, can be obtained using one of three approaches: loose mixture aging, RTFO/PAV aging, or PG, ranked with a decreasing level of accuracy. • Level 1: the log |G*| at short-term aging can be obtained by aging loose mixture in the oven at 135°C for 4 hours and then extracting, recovering, and testing the binder. The parameter M is obtained by measuring log |G*| at short-term aging and at multiple long-term aging condi- tions by aging loose mixture in an oven at 95°C and then extracting, recovering, and testing the binder. • Level 2: the log |G*| after RTFO aging can be considered equivalent to log |G*| after loose mix- ture short-term aging (4 hours at 135°C). The log |G*| values after RTFO aging and 40 hours in the PAV seem to agree with 6 days of long-term aging. The log |G*| beyond 6 days can be estimated. M can be obtained knowing log |G*| at 0 days, 6 days, and beyond 6 days.

122 Long-Term Aging of Asphalt Mixtures for Performance Testing and Prediction: Phase III Results • Level 3: the log |G*| at short-term aging and M correlate with each other and with the PG of the binder. Empirical equations can be used to convert binder PG into log |G*| at short-term aging and M. • The material-specific kinetics parameters of a mixture with a specific RAP content can be obtained from the separate virgin and RAP material-specific kinetics parameters. • The virgin binder kinetics parameters can be obtained using any one of the three levels men- tioned above. • The RAP binder kinetics parameters can be estimated from the commonly used HPG of virgin binder at the location of interest. Maps were developed to help determine the RAP HPG for a given virgin binder HPG for any location in the United States. Development of a Framework to Predict Changes in Asphalt Mixture Performance Due to Oxidative Aging AMAC, a framework that facilitates the integration of the PAM results into pavement perfor- mance prediction software, was developed by studying the changes in mixture properties with various aging levels in relation to the changes in binder properties at the same aging levels. The conclusions are as follows: • The mixture linear viscoelastic properties were measured. The mixtures exhibited an increase in modulus value, a decrease in phase angle, and a change in tTS factors with aging. • The mixture fatigue properties were measured. The fatigue analysis results show a general deterioration in performance for all the mixtures except one. The deterioration was indicated by higher and shorter damage characteristic curves, lower DR values, and lower Sapp values in comparison to the short-term aging results. • AMAC was developed to utilize the evolution of log |G*| to estimate the changes in mixture properties, namely the dynamic modulus at various temperatures and frequencies, tTS fac- tors, the damage characteristic curve, and the failure criterion, DR. • The difference between log |G*| values for short-term aging conditions and certain long-term aging conditions was used to calculate a shift factor that allows estimating the mixture prop- erties at the long-term aging condition based on known short-term aging properties. • The main required inputs for AMAC are the evolution of log |G*| at any climatic condition and depth in the pavement, which can be obtained using the PAM and the mixture short-term aging properties, whether measured or otherwise obtained. • The effects of the evolution of the material properties on structural performance were studied using FlexPAVE version 1.1. The FlexPAVE computational engine was modified to change the material properties to a stepwise incremental form as a function of aging. • The output damage results from the FlexPAVE simulations suggest that the effect of aging on pavement performance is evident only when simulations with realistic traffic and climatic conditions are considered. Suggested Research To strengthen the results of this work, more RAP field sections (with component materials and field cores preferably obtained at multiple long-term field aging durations) are needed. These sections should be obtained from different climatic conditions and be of relatively dif- ferent age levels, in contrast to the field sections included in this work that are mainly from two climatic conditions and aged for only 4 years. RAP sections are particularly complex to handle and model especially in consideration of the effects of oxidative aging. The mechanism by which mixtures containing RAP age in the field is unknown and is heavily dependent on the degree of blending between the RAP and virgin materials as well as the sources

Conclusions and Suggested Research 123   of each of the RAP and virgin materials. If 100% blending between RAP and virgin materials is assumed not true, three different kinetics reactions would be occurring simultaneously in the pavement since each of the RAP, virgin, and blended materials would have different kinetics parameters. The source of each of the RAP and virgin materials is also of importance. It is well known that virgin binders with a similar PG can have different aging susceptibilities. RAP, on the other hand, is much more complex. RAP can refer to materials that have been aged and re-aged for up to 20–30 years, and it can refer to excess materials from production with little-to-no aging. RAP can have various levels of aging depending on its previous placement in the pave- ment and the climatic conditions it was subjected to. Finally, RAP materials can have varying sources and PG of virgin binders. The work presented in this report assumed 100% blending between virgin and RAP materials. Different questions need to be tackled in future studies. One would be whether the proposed long-term aging procedure can represent the field aging condition of RAP mixtures as well as the degree of blending between RAP and virgin materials. The long-term aging procedure subjects these mixtures to 95°C for a few consecutive days. It is unknown whether 95°C can introduce more blending between virgin and RAP materials that would otherwise not exist under field conditions. The question of how to determine the kinetics parameters of RAP mixtures would need to be thoroughly investigated. Extraction and recovery of a RAP mixture impose 100% blend- ing. Should virgin and RAP materials then be aged separately? This report provided a means by which the kinetics parameters of the blended mixture can be obtained when the kinetics parameters of the virgin and RAP materials are known. Is it a feasible idea to age RAP materials separately? Can the RAP kinetics parameters be predicted? This report showed that the PAM can potentially be used to estimate the RAP kinetics parameters. The PAM can predict the evolu- tion of binder properties given the pavement temperature history and the initial binder kinetics. With some assumptions for initial binder kinetics, RAP properties can be estimated for any location and climatic condition. Thus, a map was generated where the RAP HPG can be picked out for any location of interest. Picking out the RAP properties based on a map for any loca- tion can be extremely useful in obtaining the aging kinetics of RAP mixtures. While all these tools are promising, more RAP materials, mixtures, and field sections are needed to thoroughly investigate the effect of RAP blending, the age level of RAP, sources of RAP, and RAP content.