Relationship Between Chemical Makeup of Binders and Engineering Performance (2017)

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.

53 CHAPTER FIVE CASE EXAMPLES OF BINDER CHARACTERIZATION PRACTICES INTRODUCTION Based on the survey responses, a number of agencies indicated that they are in the process of implementing performance specifications for asphalt mixtures or conducting research with performance tests to support their eventual implementation. There were a number of reasons why these agencies were selected to serve as case examples. They represent a geographic distribution that includes a wide range of climates and paving program sizes. The responses from the agency survey indicated that Louisiana, Virginia, and Missouri in particular have made considerable advancements in binder evaluation. This chapter discusses case examples from three agencies in the United States (Louisiana, Virginia, and Missouri DOTs) and one Canadian Ministry of Transportation (Ontario). Their states of practice were captured through interviews with one or multiple members of each agency to determine (1) how performance specifications for asphalt binders were developed and are used, (2) the evaluation of performance testing, and (3) implementation efforts related to performance specifications for asphalt mixtures. LOUISIANA Lousiana Department of Transportation and Development (LDOTD) reported that asphalt mixture designs are approved in the nine LDOTD district offices, and the responsibility to let pavement overlay and road contracts is also at the district level. There are five major asphalt producers, and 15 to 20 contractors overall, that undertake state paving jobs. Four of the produc- ers modify crude oil residuum and their products vary according to crude oil source. One of the producers blends asphalt exclusively and ships a very consistent product that is more compatible with polymeric additives. All contractor laboratories (in trailers or buildings) must be accredited through the Construction Materials Engineering Council or AASHTO Materials Reference Laboratory. The state listing generally comprises 30 asphalt plants, and most contrac- tors select a fully equipped central location to do their initial mix design testing, which does include some performance tests. At this time, the state does not employ a consultant-based system and, as a result, LDOTD noted that it has been a challenge to implement performance tests such as those conducted with AMPT equipment, owing to the cost and complexity of these tests. Development and Use of Binder Specifications The LDOTD Materials Laboratory has revised the binder specifications based on several years of experimentation with dif- ferent testing techniques. The current protocol measures rotational viscosity, dynamic shear (DSR) G*/Sin Delta and phase angle at 10 rd/s, flash point solubility, and polymer separation on the original binder. The rotational viscosity is measured to determine product uniformity by the supplier. A binder having a rotational viscosity of 3.0 Pa·s or less will typically have adequate mixing and pumping capabilities. The samples for the separation are prepared according to ASTM D7173. The softening point is determined at the top and bottom of the tube per AASHTO T 53. The separation test is not required when crumb rubber is used. Tests on RTFO residue include mass change, DSR G*/Sin Delta, elastic recovery, ductility, and multiple stress creep recov- ery. Further testing of PAV residue is completed using DSR and bending beam. Issues with the force ductility test have led to the replacement of this test with an initial DSR phase angle maximum of 75°@ 76°C for polymer-modified asphalts. The force ductility tests indicated the presence but not the quality of the polymer modifiers. All the base asphalts used in Louisiana meet PG 67-22 specifications. A maximum of 10% crumb rubber is allowed in PF 82-22m and PG 76-22m mixes. Samples meeting the specifications listed in Table 13 for polymer-modified grades are verified for acceptance; true grading of the sample is not done.

54 TABLE 13 LOUISIANA PERFORANCE GRADED ASPHALT CEMENTS Property AASHTO Test Method PG 82-22rm PG 76-22m PG 70-22m PG 67-22 PG 58-28 Spec. Spec. Spec. Spec. Spec. Tests on original binder: Rotational viscosity @ 135°C, Pa·s T 316 3.0 3.0 3.0 3.0 3.0 Dynamic shear, 10 rad/s, G*sin delta, kPa T 315 1.00+ @ 82°C 1.00+ @ 76°C 1.00+ @ 70°C 1.00+ @ 67°C 1.00+ @ 58°C Dynamic shear, 10 rad/s, phase Aangle, ° T 315 — 75° @ 76°C — — — Flash point, °C T 48 232+ 232+ 232+ 232+ 232+ Solubility, % T 44 N/A 99.0+ 99.0+ 99.0+ 99.0+ Separation of polymer, 163°C, 48 hours, degree C difference in R & B from top to bottom ASTM D7173 AASHTO T 53 — 2- 2- — — Tests on rolling thin-film oven residue: T 240 Mass change, % T 240 1.00- 1.00- 1.00- 1.00- 1.00- Dynamic shear, 10 rad/s, G*/sin delta, kPa T 315 2.20+ @ 82°C — — 2.20+ @ 67°C 2.20+ @ 58°C Elastic recovery, 25°C, 10 cm elongation, % T 301 60+ — — — — Multiple stress creep recovery (MSCR), 67ºC, Jnr(3.2 kPa) T 350 — 0.5- 2.0- — — Multiple stress creep recovery (MSCR), 67ºC, % recovery (3.2 kPa) T 350 — Meets curve Meets curve — — Ductility, 25°C, 5 cm/min, cm T 51 — — — 90+ — Tests on pressure aging vessel residue: R 28 Dynamic shear @ 26.5°C, 10 rad/s, G* sin delta, kPa T 315 5,000- 6,000- 6,000- 5,000- 5000- @ 19°C Bending beam creep stiffness, S, MPa @ –12°C. T 313 300- 300- 300- 300- 300- @ –18°C Bending beam creep slope, m value @ –12°C T 313 0.300+ 0.300+ 0.300+ 0.300+ 0.300+ @ –18°C Source: Louisiana DOTD (2016). The MSCR curves and Jnr limits as defined by AASHTO M 332 are effective monitors of binder quality. Some of the plant- blending operators can meet the Jnr, but they have had issues meeting the curve for PG 70-22m and PG 76-22m. They try to balance stiffening the binder enough to meet the initial DSR phase angle maximum of 75°C (an indicator for the presence of polymer) but not stiffen it so much as to cause issues with the recovery. For the PG 70-22m, the problem appears to be similar in that the blenders are trying to meet the recovery curve by adding more polymer (latex, generally), which stiffens the asphalt and raises the recovery target. Some are now using additives that appear to help, and using a different base asphalt appears to make a difference. MSCR percent recovery identifies elastomeric polymer, but MSCR recovery curves were developed using SBS modifiers. Additional curves for SBR, crumb rubber, and so forth are needed to specify a percent recovery for a specific Jnr. These curves are even more important if the actual Jnr is used rather than a maximum for a grade. GPC Analysis The Louisiana Transportation Research Center (LTRC) is the development arm for new techniques. LTRC developed a method using deconvoluted GPC chromatograms to determine polymer percent in modified asphalt and examine asphal- tene/maltene ratios. A robust state-of-the art GPC system was installed in an LDOTD Materials Laboratory room dedicated

55 solely to preparation and analysis of samples received from Louisiana suppliers of paving asphalt materials. An effective asphalt binder extraction methodology without affecting the binder properties was developed. Both virgin asphalt cements and asphalt concrete cores from field projects are analyzed. Asphalts supplied by each individual producer yield specific GPC curves that confirm the specified quantities of the polymer added and serve as standards for forensic analysis. Asphalt binders extracted from field cores yield the curves characteristic of the asphalt producer. The library of GPC curves has already been used effectively to identify the source of asphalts used in the field. Plans to require each project to submit an acceptance core from the first lot are being considered. Future research on the application GPC to asphalt binder changes after field aging is planned. LDOTD noted that there have not been many instances of rutting in existing Superpave asphalt pavements, except for some extraneous reasons, such as issues at the plant during production, issues in the field during construction, or material problems. Therefore, it indicated that the control of moisture damage and fatigue were performance-based tests the agency considers most important for predicting the pavement performance typical for its roadway system. All mixtures are required to contain at least 0.6% antistripping agent, which is added at the plant. A maximum of 20% RAP is allowed in the mixtures. Recycling agents may be added to mixtures as long as all the speci- fications for the mixtures are satisfied. There is a moratorium on the use of RAS in Louisiana. Future research reported by LDOTD also includes its involvement in a Pooled Fund Study TPF-5(294) being conducted by the LTRC that will evaluate various fatigue tests, including identifying differences in how the results from these tests differ for mixtures produced with higher RAP or RAS contents. Specifically, one objective of the research project is to establish mechanistic test criteria for achieving durable flexible pavements made from both WMA and HMA mixtures that contain high RAP and/or RAS con- tents. A second objective is to develop preliminary asphalt mixture specifications that incorporate the resulting mechanistic test criteria to be tested on plant-produced specimens and/or roadway cores, as based on the results of the study. The testing of plant-produced mixtures and roadway cores will facilitate the evaluation of the impacts of higher RAP and/or RAS per- centages on the durability of the asphalt mixtures investigated as part of the study. VIRGINIA Virginia has the third largest state-maintained highway system in the United States, with more than 57,000 miles of roadways, the vast majority of which are constructed using asphalt. Virginia Department of Transportation (VDOT) reported that the approval of asphalt mixture designs occurs at the district level in each of nine districts. The responsibility to let pavement contracts also exists within the district offices. There are 27 asphalt producers in Virginia. Most contractors design their mix- tures at a central laboratory. Contractor labs either receive AASHTO accreditation or are certified through VDOT inspection, which uses a protocol similar to AASHTO’s. Development and Use of Binder Specifications VDOT adopted Superpave binder performance grading in the mid-1990s and in 1997 began to using the Superpave mix design system. Since adopting the PG system, no major changes have been made to VDOT’s specifications. In 2015, VDOT adopted the multiple stress creep recovery test. This test is more efficient than the Superpave elastic recovery test, and makes identification of polymer modification in binders easier (Table 14). Virginia is identified as a “PG 64 state,” so all MSCR testing is conducted at 64°C. The designations after the testing temperature (64°C unless otherwise specified) generally indicate the effective change in high temperature grade to accommodate expected traffic loading. These letter designations are “S” for standard, “H” for heavy, “V” for very heavy, and “E” for extra heavy. The new MSCR binder designations and their corresponding Superpave PG designations are listed in Table 1. Specific VDOT binder testing require- ments are provided in Tables 15 and 16. Asphalt pavements are subject to density requirements, operating within mix design tolerances for in-place density (as measured by nuclear gauge), asphalt content, gradation, voids total in the mix, voids in the mineral aggregate, voids filled with asphalt, and dust-to-asphalt-cement ratio. Asphalt content and gradation are checked for every 500 tons of mix. Volumetrics are tested for every 1,000 tons of mix. Asphalt manufacturers typically take quality control (QC) samples of binder for every batch, and VDOT does verification/independent assurance.

56 TABLE 14 MULTIPLE STRESS CREEP AND RECOVERY VERSUS PERFORMANCE GRADE DESIGNATIONS MSCR Designation Superpave PG Designation PG 58S-28 PG 58-28 PG 64S-22 PG 64-22 PG 64H-22 PG 70-22 PG 64V-28 PG 70-28 PG 64E-22 PG 76-22 Source: Virginia DOT (2016). TABLE 15 VDOT SPECIFICATIONS, REFERENCE AASHTO M-320 Property AASHTO Test Method PG 64E-22 (PG 76-22) PG6 4H-22 (PG 70-22) PG 64S-22 (PG 64-22) PG 58S-28 (PG 58-28) Spec. Spec. Spec. Spec. Tests on original binder: Rotational viscosity @ 135°C, Pa·s T 316 3.0 3.0 3.0 3.0 Dynamic shear, 10 rad/s, G*/sin delta, kPa T 315 1.00+ @ 76°C 1.00+ @ 70°C 1.00+ @ 64°C 1.00+ @ 58°C Dynamic shear, 10 rad/s, phase angle, ° T 315 — — — — Flash Point, °C T 48 230+ 230+ 230+ 230+ Solubility, % T 44 99.0+ 99.0+ 99.0+ 99.0+ Separation of polymer, 163°C, 48 hours, degree C difference in R & B from top to bottom ASTM D7173 AASHTO T 53 2- 2- — — Tests on rolling thin-film oven residue: T 240 Mass change, % T 240 1.00- 1.00- 1.00- 1.00- Dynamic shear, 10 rad/s, G*/sin delta, kPa T 315 2.20+ @76 2.20+ @70 2.20+ @ 64°C 2.20+ @ 58°C Elastic recovery, 25°C, 10 cm elongation, % T 301 — — — — Ductility, 25°C, 5 cm/min, cm T 51 — — 90+ — Tests on pressure aging vessel residue: Dynamic shear, @ 26.5°C, 10 rad/s, G* Sin Delta, kPa T 315 5000- @ 31°C 5000- @ 28°C 5000- @ 25°C 5000- @ 19°C Bending beam creep stiffness, S, MPa @ –12°C. T 313 300- 300- 300- 300- @ –18°C Bending beam creep slope, m value,@ –12°C T 313 0.300+ 0.300+ 0.300+ 0.300+ @ –18°C Source: Virginia DOT (2016). Binder Modification and/or Use of Reclaimed Material VDOT permits up to 30% RAP in asphalt mixtures using a binder performance grade PG 64S-22 (PG 64-22) base. RAS have a 5% allowance, but the binder must meet PG 64H-16 (PG 70-16) and must use PG 64S-22 (PG 64-22) as the base mix binder grade. Rubber-modified binders are not prohibited by VDOT as long as they pass current binder specifications. There is not a separate specification for rubber-modified binders, as they have received limited use. There are no specified polymer modifiers; however, all modified binders are subject to the same binder testing (e.g., PG binder grading and MSCR testing). In-line blending of polymers is very limited in Virginia and is subject to approval. Currently, most in-line blending is used only for special projects. VDOT does allow warm-mix asphalt to be used for paving projects. Although some projects use warm-mix additives to produce WMA, many contractors also use additives or foaming technologies as a compaction aid for HMA or to place asphalt at a production temperature between WMA and HMA. All additives are subject to product approval by VDOT.

57 TABLE 16 VDOT SPECIFICATIONS, REFERENCE AASHTO M-332 Property AASHTO Test Method PG 64E-22 (PG 76-22) PG 64H-22 (PG 70-22) PG 64S-22 (PG 64-22) PG5 8S-28 (PG 58-28) Spec. Spec. Spec. Spec. Tests on original binder: Rotational viscosity @ 135°C, Pa·s T 316 3.0 3.0 3.0 3.0 Dynamic shear, 10 rad/s, G*/sin delta, kPa T 315 1.00+ @ 76°C 1.00+ @ 70°C 1.00+ @ 64°C 1.00+ @ 58°C Dynamic shear, 10 rad/s, phase angle, ° T 315 — — — — Flash point, °C T 48 230+ 230+ 230+ 230+ Solubility, % T 44 99.0+ 99.0+ 99.0+ 99.0+ Tests on rolling thin-film oven residue: T 240 Mass change, % T 240 1.00- 1.00- 1.00- 1.00- Elastic recovery, 25°C, 10 cm elongation, % T 301 — — — — Multiple stress creep recovery (MSCR), 64ºC, Jnr (3.2 kPa) T 350 @64°C @ 64°C @ 64°C @ 58°C Multiple stress creep recovery (MSCR), 64ºC, % recovery (3.2 kPa) T 350 > 29.371 — — — Tests on pressure aging vessel residue: R 28 Dynamic shear, @ 26.5°C, 10 rad/s, G* sin delta, kPa T 315 6,000 6,000 5,000- 5,000- @ 19°C Bending beam creep stiffness, S, MPa @ –12°C. T 313 300- 300- 300- 300- @ –18°C Bending beam creep slope, m value @ –12°C T 313 0.300+ 0.300+ 0.300+ 0.300+ @ –18°C Source: Virginia DOT (2016). Specialty Testing The Virginia Transportation Research Council is the research arm of VDOT. The council is often tasked with aiding the VDOT Materials Office with evaluating potential changes to specifications and test procedures. It also leads research and development projects that examine new test methodologies. Past and current research projects examined different nontradi- tional techniques for testing asphalt binders. Among these are Fourier FTIR, GPC, and XRF. In 2006, research was completed that used FTIR as a method to identify the presence of polymer modifiers in asphalt bind- ers. This method was found to identify the presence of polymers with no false-positives and could be calibrated to quantify the amount of modifier present based on polymer type and base binder source. Calibration, however, would require extensive effort and was deemed to not be practical. Thus, FTIR remains more qualitative than quantitative for identifying polymer modification. FTIR was also found to be useful in identifying whether asphalt binder solvents were completely removed from the binder after recovery. VDOT’s use of GPC is primarily in the investigative stage. Its primary use, thus far, is to examine molecular changes caused by binder oxidation. Potential future uses involve forensic analysis, investigation of polymer degradation with age, and development or implementation of quantitative methods for determining polymer content. There is also interest in utilizing XRF to determine binder composition, specifically to identify the presence of refined engine oil bottoms and other additives. MISSOURI Missouri allows antistripping agents and wax-based additives (warm mix) in its mixture designs. Binder samples are taken when they contain all of the additives, that is, just before the binder enters the mixing drum. Daily binder samples are required. If the selected samples do not meet the standards, more extensive testing is employed before the material will be accepted.

58 Missouri focuses on the end result. Randomly chosen field samples are checked using DSR to confirm the PG binder grade. Acceptable tensile strength ratio data (quality control and quality assurance) are required on field samples for acceptance; this is a check on the presence of antistripping agents. Occasionally the Missouri DOT observed that some antistripping agent can affect the binder grade. The contractor may use whatever amount of antistripping agent is necessary to meet the TSR. Missouri has one binder supplier that utilizes PPA. The supplier must indicate the use of PPA on the bill of lading and advise the asphalt paving contractor of its presence so the PPA can be incorporated in the mix design. In a few jobs where the asphalt paving contractor was not aware of the presence of PPA, the PPA neutralized the antistripping agent, which negatively affected the field TSR. At least one antistripping agent that is compatible with PPA has been identified. The PPA also reacted with the polymer and negatively affected the PG binder grade. To eliminate the possibility of further problems, the Missouri DOT has developed a test method for determining the amount of PPA in PG-grade asphalt binders. PG-grade asphalt binder is fused with sodium carbonate and dissolved in an acid solution. The solution is analyzed for phosphorous on an inductively coupled plasma optical emission spectrophotometer and the phosphorus content is calculated as percent of PPA. This test is used when PG binder grade is not confirmed through DSR testing. Sustainability Efforts Missouri has been incorporating RAP into asphalt cements since the 1990s. Beginning in 2013 95% of HMAs contained recycled materials. An average of 23.8% RAP and 2.5% RAS is present in Superpave mixes. Stone-mastic asphalts have been used on an experimental basis only. Contract-grade virgin binder may be used to blend up to 20 weight percent RAP or up to 10 weight percent RAS. The contribution of the RAP to the binder content is estimated as an effective virgin binder replacement percentage. If an RAP/RAS combination is present at less than 20 weight percent, the effective virgin binder replacement estimated as [RAP + (2*RAS)] may be added to the contract grade virgin binder as long as the final binder blend meets the designated PG specifications. If up to 40% RAP is blended, the PG of the virgin binder must be reduced one grade. The standard PG 64-22 binder may be modified with recycling (softening) agents to produce a PG 58-28 binder grade before the RAP addition. Specific blend charts are required for cements containing greater than 40 weight percent RAP. The incor- poration of RAS began in 2009; initially up to 8% was added but the amount has now been reduced to 2%–3%. If producers use recycling agents, the dosage rates and the volumetrics must be designated in the mix design. Acceptance of the blended binders depends on meeting the designated performance grade specifications. Ground tire rubber may also be employed if the producers do terminal blending and provide an acceptable QC plan. Up to 8% GTR may be added to PG 64-22 base binder to produce a PG 76-22 binder. All blends containing GTR include 4.5% transpolyoctenamer rubber by weight of the GTR. The direct tension test is waived, but a separation test is performed in accordance with ASTM D 5976. The number of Superpave mixes containing GTR was as high as 22% in 2011, but currently approximately 10% of the designs include GTR. ONTARIO The Ontario Ministry of Transportation (MOT) is one of the largest in Canada, with more than 16,500 km (~10,300 m) of roadways, the vast majority of which are constructed using asphalt. Ontario is Canada’s most populous province by a large margin, accounting for nearly 40% of all Canadians, and is the second-largest province in total area. The surrounding Great Lakes greatly influence the climatic region of southern Ontario. Proximity to the Great Lakes gives some parts of southern Ontario milder winters. The climate is similar to that of the inland Mid-Atlantic states and the Great Lakes portion of the Midwestern United States. The region has warm to hot, humid summers and cold winters. The northernmost parts of Ontario, primarily north of 50°N, have a subarctic climate with long, severely cold winters and short, cool to warm summers with dramatic temperature changes possible in all seasons. Temperatures of −40°C (−40°F) are not uncommon; snowfall remains on the ground for sometimes over half the year. Thus, MOT must cope with extreme environments in designing highways and specifying the materials for their construction. Five MOT regional offices are responsible for the planning, design, construc- tion, maintenance, operations, and management of the provincial highway network. In the 1990s, the Strategic Highway Research Program Performance Graded Asphalt Cement specifications were imple- mented with the expectation that identifying asphalt cements that meet the specifications would help to reduce pavement cracking. Over the past 15 years, MOT has observed that some pavements experience more cracking within 2 to 3 years of placement, whereas others experience little to no cracking in their early life. Through numerous trial contracts and exten- sive research, MOT identified poor-quality asphalt cement as one of the primary causes for premature pavement cracking.

59 Chemical analysis showed that the likely presence of waste engine oil residues, air-blown residues, and/or acids in some of the asphalt cements may have contributed to excessive cracking. The acceleration of the physical hardening phenomena using an isothermal conditioning is correlated to data from seven trial sections and regular contracts (Evans et al. 2011). Tensile specimens are poured and conditioned for either 20 min or 72 h at low temperatures before being subjected to a stress relaxation test at −10°C. The extended conditioning period could more than double the residual thermal stress at the end of the test. The one asphalt to show no physical hardening came from the section that remained largely free of cracking. A second material that showed a moderate degree of physical hardening only recently started to crack by an appreciable amount. In contrast, the remaining five materials significantly hardened during the extended conditioning period and cracked prematurely and excessively in service. Materials with an unstable colloidal structure were found most sensitive to physical hardening. The results of the study agree with earlier creep data obtained according to an extended bending beam rheometer test method (Hesp et al. 2009). Hence, it is imperative that pavements be designed with criteria that take physical hardening effects into account to limit premature and excessive performance failures. Analysis of Binder Crack Resistance In the past decade one lab noticed premature cracking (top-down mostly), starting in the wheel paths and then propagating in a map cracking form (Tabib et al. 2015). The lab recovered the binder in several cases and noticed a great loss of low tem- perature grade using an extended bending beam rheometer (ExBBR) test. This is an indication that the SHRP aging protocol is not sufficient on the low temperature side. The research on binders extracted from poorly performing pavements led MOT to focus on two of these test methods to better predict premature pavement cracking. Descriptions of these two tests follow. ExBBR, which determines the low-temperature continuous performance grade of physically hardened asphalt cement using different conditioning temperatures and times than those used in the AASHTO T 313 test method for bending beam rheometer. The BBR method conditions samples for 1 h and tests them at a temperature 10ºC warmer than the low tempera- ture grade (T). The ExBBR method conditions samples up to 72 h at 10ºC warmer (T +10ºC) and 20ºC warmer (T + 20ºC) to determine, by extrapolation, the limiting temperature for each conditioning time and temperature. Double-edge notched tension (DENT) is an indicator of asphalt cement’s resistance to ductile failure. The test is conducted after intermediate thermal conditioning to determine the essential work of fracture, the plastic work of fracture, and an approximate critical crack tip opening displacement at a specified temperature and rate of loading. The ExBBR test is intended to predict reversible physical hardening that can cause premature low-temperature cracking. The DENT test is used as a measure of asphalt cement’s elasticity, or ability to stretch and resist cracking at intermediate temperatures. Because the tests look at different thermal states of the pavement, and strain tolerances change with tempera- ture as a result of variations in hardening tendency, it is necessary to review both ExBBR and DENT results to accurately predict the pavement cracking performance (Hesp et al. 2009; Shurvell et al. 2009). Hesp et al. (2009) found that factors such as chemical and physical hardening cannot be captured by a single test at 15°C and that the DENT test is intended to exclude poor performers rather than to provide a perfect correlation with pavement cracking [5]. The criteria for passing these tests has been incorporated into MOT’s specification for performance graded asphalt cement acceptance (Table 17). Ontario MOT requires the use of antistripping additives for northern Ontario and allows a limited quantity of PPA. Because much PG XX-34 is used and some PG XX-40, MOT knows that softening agents (oils, etc.) are being used, but these are not specified. The final product must comply with M320 specifications. Efforts to use XRF to detect PPA failed to give accurate results owing to low dosage. MOT tried using FTIR to analyze for antistripping agents, but the low dosage is hard to detect. Ongoing research includes the use of XRF and FTIR to quantify REOB, which currently is limited by the ash content speci- fication. A modified PAV procedure is being evaluated, which would look at either reducing the asphalt film thickness or increasing the oven aging time. SUMMARY The interviewees from DOTs of three states and one Canadian MOT consistently observed the following points: • The Strategic Highway Research Program Performance Graded Asphalt Cement specifications have deficiencies. • Estimation of low-temperature performance can be improved using combined ExBBR and DENT fatigue testing.

60 • The MSCR test is more efficient compared with the Superpave elastic recovery test, and makes identification of polymer modification in binders easier. • Greater use of recycled materials (e.g., RAP, RAS, and crumb rubber) required adjusting the AC content to ensure suf- ficient liquid in the mixes. • The interviewed agencies are moving toward incorporating performance testing for production acceptance; however, the barriers to more widespread adoption of performance tests are the cost, manpower, and time required to run the performance tests. Tests under extensive review include XRF, FTIR, and GPC. TABLE 17 ONTARIO MINISTRY OF TRANSPORTATION SPECIFICATION FOR PERFORMANCE-GRADED ASPHALT CEMENT ACCEPTANCE Property and Attributes (Unit) Test Method Acceptable Major Borderline Rejectable PGAC Grade Ash content, % by mass of residue (%) LS-227 ≤ 0.8 > 0.8 and ≤ 1.0 > 1.0 All PGAC except PG 52-40 and PG 58-40 ≤ 1.0 N/A > 1.0 PG 52-40 and PG 58-40 Low-temperature limiting grade (°C) LS-308 < -YY (low PGAC grade) 3 to 6°C warmer than low PGAC grade > 6°C warmer than low PGAC grade All PGAC Grades except PG 58-28 and PG 52-34 Grade loss (°C) LS-308 0–6 6–8 > 8 Nonrecoverable creep compliance at 3.2 kPa (Jnr3.2) (kPa-1 ) AASHTO T 350 testing conducted at 58°C in southern Ontario and 52°C in northern Ontario < 4.5 N/A ≥ 4.5 Average percent recovery at 3.2 kPa (R3.2) (%) > the lesser of [(29.371) (Jnr3.2)- 0.2633] or 55 N/A ≤ the lesser of [(29.371) (Jnr3.2)- 0.2633 -10] or 45 Percent difference in nonrecoverable creep compliance between 0.1 kPa and 3.2 kPa, Jnrdiff (%) Testing carried out only for information purposes CTODs, delta (mm) LS-299 ≥ 10.0 mm for PGAC XX-28 < 6 and ≥ 4 mm for PGAC XX-28 < 4 mm ≤ 14.00 mm for PGAC XX-34 < 10 and ≥ 8 mm for PGAC XX-28 < 8 mm ≥ 18.0 mm for PGAC XX-40 < 14 and ≥ 12 mm for PGAC XX-28 < 12 mm Source: Ontario MOT (2016).