Investigation of Low and High Temperature Properties of Plant Produced Rap Mixtures

RESEARCH APPROACH

This section describes the approach used in both phase I and phase II of the current study to evaluate the effects of RAP on the properties of plant-produced asphalt mixtures. The results from phase I are summarized in the next section and full details are provided in the final unpublished report. Where appropriate, the results from testing mixes produced by the phase I contractor are combined with those of the phase II contractors for completeness.

Mix Design, Production, and Sampling

In phase II of this study, four contractors replicated the experiment conducted by one contractor in phase I of this experiment. Each contractor designed six mixtures, as shown in table 1, to be as similar as possible. The mixes are SUperior PERforming Asphalt PAVEments (Superpave^®) mixes with a nominal maximum aggregate size (NMAS) of 0.37 inches ³/₈ inches (9.5 mm). Each contractor’s set of mixes used one source of RAP, one set of virgin aggregates, and one source of each of two binder grades. That is, each binder grade was from one source per plant, although the different binder grades may have come from different sources. The RAP contents ranged from zero (the control mix) to 40 percent, as shown in table 1.
AMENDED 1/10/2012

The contractors generally did a complete mix design on one mixture and then altered the binder grade or aggregate/RAP stockpile percentages to fill the other cells. These alternate mixes were typically verified to conform substantially to the mix design requirements with a one-point mix design. The mix designs and/or quality control (QC) testing results provided by the contractors are in appendix A of this report.

In almost all cases, the mix design gradations for a given contractor agreed within a range of 3 percent or less on each sieve. Consequently, variations in the voids in the mineral aggregate (VMA) and sometimes total binder content occurred as the aggregate properties changed when RAP was added to the mixtures. One contractor used exactly the same mix design gradation for all six mixes and varied the binder content to achieve 4 percent air voids (AV) at design. The greatest differences in the gradations were for contractor 1. Those mix designs agreed within less than 3 percent except on the 0.37- ³/₈-inch (9.5-mm) 0.09-, 0.05-, and 0.02-inch (9.5-, 2.36-, 1.18-, and 0.6-mm) No. 8, No. 16, and No. 30 sieve. None of the mixes from contractor 1 were consistently higher or lower than the others, even where the greatest ranges occurred. That is, the higher RAP contents did not necessarily yield finer mixes or vice versa. Contractors 1 and 4 provided coarse mix designs, whereas contractors 2, 3, and 5 offered fine mixes (compared to the primary control sieve control point of 46 percent passing the 0.09-inch (2.36-mm) No. 8 sieve).⁽⁴⁾
AMENDED 1/10/2012

Two different binder grades, PG58-28 and PG64-22, were used. PG64-22 is the standard binder for the area, and PG58-28 is the grade that would be selected for the 25 percent RAP content mixtures. For the 40 percent RAP content mixtures, a blending chart would be required to determine the appropriate virgin binder grade according to the current AASHTO standards.⁽⁴⁾

The mixes were produced through the contractors’ hot mix plants (over as short a time frame as practical) using any processing they typically use with RAP mixes. It was requested that approximately 100 T (90.7 90 Mg) of each mixture be produced before sampling. The contractors placed the mixes wherever they could, typically on commercial or local road projects.
AMENDED 1/10/2012

The contractors sampled the mixes from a truck at the plant and stored the samples in sealed 5-gal (18.93 20-L) buckets. The contractors also sampled the RAP stockpile and virgin binder. The following minimum samples were requested:

• RAP: Three 5-gal (18.93 20-L) buckets.
  
  
• Loose mix: Eight 5-gal (18.93 20-L) buckets per mix.
  
  
• Liquid: Two 1-gal (18.93 4-L) paint cans of each grade of asphalt binder
  
  
• Compacted samples: Three field-mixed plant lab-compacted gyratory samples per mix.
  AMENDED 1/10/2012

The contractors were asked to provide information on the maximum theoretical specific gravity of each mix, binder content, gradation, plant type, tonnage, and any RAP processing techniques used. In some cases, this level of detail was not provided.

The details of the contractors who agreed to participate in this study and information about their plants are provided in table 2. RAP from Michigan would have been produced originally using a softer binder grade than the Indiana sources because of the climate and prevailing specifications.

Laboratory Testing Plan, Test Procedures, and Data Analysis

The laboratory testing plan was designed to examine the following:

• Verification of mixture volumetrics and compliance with applicable standards (including binder grade).
  
  
• Assessment of the high- and low-temperature mixture properties and the effect of increasing RAP content on those properties.
  
  
• Examination of the degree of blending of the RAP and virgin binders in the plant-produced mixtures.

The following tests were conducted on various samples:

• Volumetric data on plant-produced and compacted mixture samples. The maximum specific gravity (G_mm) was measured on two samples of each mixture according to AASHTO T 209, and the results were averaged for AV content determination.⁽¹³⁾ The bulk specific gravity (G_mb) was measured according to AASHTO T 166.⁽¹⁴⁾ AV content was determined using the measured maximum theoretical and G_mb of the mixtures. The G_mb and AV content of all laboratory compacted samples was also determined to ensure compliance with the sample requirements for the mechanical tests being conducted. Compaction of the specimens was conducted according to AASHTO T 312.⁽¹⁵⁾
  
  
• Verification of virgin binder grade. Tank samples of each virgin binder were tested for compliance with the AASHTO M 320 high- and low-temperature grades.⁽¹⁶⁾ Frequency sweeps and temperature sweeps using the dynamic shear rheometer (DSR) were conducted on the original binder to determine the complex shear modulus (|G*|_m).⁽¹⁷⁾ Two replicates were tested for each binder and averaged to develop the master curves for the binders. Three replicates were tested at low temperatures in the bending beam rheometer (BBR) after rolling thin film oven (RTFO) and pressure-aging vessel aging.(18–20) These data were used to determine the true (or continuous) grade of the binder.
  
  
• Determination of mixture properties. Lab-compacted mixture samples were tested for their high- and low-temperature properties. To do so, 5-gal (18.93 20-L) buckets of the plant-produced mix were heated to approximately 239 240 °F (115 °C) for typically 1 h, which was just long enough to soften the mix for splitting. Samples of the loose mix were conditioned for 2 h at the compaction temperature (289.4 290 °F (143 °C)) before compaction into gyratory specimens. The gyratory specimens were then cut to the proper size for the specific mix tests to be performed.
  AMENDED 1/10/2012
  
• Four replicates for each mix (compacted to 7 ±0.5 percent AV) were tested for |E*| using the universal testing machine (UTM-25) developed by Industrial Process Controls, Ltd. Samples were tested at 39.2, 69.8, 98.6, and 129.9 40, 70, 100, and 130 °F (4, 21, 37, and 54.4 °C) at the frequency range specified in AASHTO TP 62.(21) These results were then used to develop master curves in Microsoft Excel^®, which show the trends in |E*| as a function of reduced frequency. The modulus of the mixtures can be related to pavement rutting at high temperatures and fatigue cracking at intermediate temperatures.
  AMENDED 1/10/2012
  
• The mixes were also tested at low temperatures for indirect tensile (IDT) creep compliance and strength (specified in AASHTO T 322) to estimate the effects of RAP content and properties on the thermal cracking behavior of the resulting mix.⁽²²⁾ Creep compliance tests were conducted for 100 s on three replicates (compacted to 7 ±0.5 percent AV) at -4, 14, and 32 °F(‑20, ‑10, and 0 °C). Following the creep compliance tests, the same samples were tested for IDT strength at 14 °F (-10 °C). These data were used to determine the critical cracking temperatures (T_crit) of the mixes.
  
  
• Extraction and recovery of binder. The binder from each RAP source and mixture was extracted and recovered according to AASHTO T 319  using n-propyl bromide (nPB).⁽²³⁾ The recoveries in phase I were conducted using methylene chloride (mCl) and the Abson procedure, which is discussed later in this report.⁽²⁴⁾ The recovered binders were then tested for complex modulus (|G*|) in the DSR.⁽¹⁷⁾ This information was used according to Bonaquist’s method of evaluating binder blending.⁽¹¹⁾ The recovered binders were also tested at low temperatures as described above for virgin binders (except without RTFO aging since they had been through a hot mix plant).
  
  
• Comparison of extraction/recovery techniques. Samples of selected mixtures were also extracted according to AASHTO T 164 and recovered using the Abson method with mCl, as done in phase I of this study.^(25,24) They were also extracted with nPB for comparison to the AASHTO T 319 results.⁽²³⁾ The Abson recovered binders were then tested in the DSR, and the properties were compared to those of binders recovered according to AASHTO T 319. This was done to assess the impact of the extraction/recovery technique/solvent on the resulting binder properties. The Abson recovery technique was used in phase I of this study and may have contributed to the observed results.⁽²⁴⁾

Collaborations

The research team provided samples of one contractor’s mixtures (set of six) to the FHWA Turner-Fairbank Highway Research Center (TFHRC) for study. Researchers at TFHRC performed fatigue testing on these materials, utilizing a pull-pull fatigue test as a part of their research. Those results are summarized later in this report.

Samples were also provided to Dr. Hussain Bahia from the University of Wisconsin—Madison University of Wisconsin–Madison, for his use in a RAP mortar testing procedure. Those results will be reported by Dr. Bahia separately as part of his overall project. |E*| and binder testing data were shared with Dr. Jo Daniel at the University of New Hampshire as she developed plans for a pooled fund study of plant-produced RAP mixtures in the Northeast. Lastly, samples of binders recovered from one set of mixes were provided to Dr. Eric Kalberer at the Western Research Institute for Atomic Force Microscopy compatibility testing.
AMENDED 1/11/2012