Using Recycled Asphalt Shingles with Warm Mix Asphalt Technologies (2018)

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.

124 Mix Design Verifications The mixtures from the new projects (Wisconsin, Alabama, Tennessee, North Carolina, and Indiana) were verified accor ding to AASHTO PP 78-17. For each project, the cold-feed percentages from plant production were used to reproduce the as-produced gradations as much as pos- sible. In addition, three to four asphalt contents following the Super pave volumetric methodology were used to verify mixes for the Wisconsin, Alabama, North Carolina, and Indiana projects. Four asphalt contents were used for the Tennessee mix (Marshall Mix Design Method). To assess compliance with the most recent version of AASHTO PP 78-17, binders were recovered from the mix design verification samples at the optimum asphalt content and conditioned for 40 h in the PAV before testing in the bending beam rheometer so that the DTc parameter could be determined in accordance with the new criterion. The 2017 version of AASHTO PP 78 also requires increasing the minimum VMA of mixtures containing RAS compared to the requirements in AASHTO M 323. This change was made to account for RAS binder that does not activate during the mixing process. For mixtures containing RAS, the minimum VMA is increased by 0.1% for one percent of RAS by weight of total aggregate. One goal of the mix design verifications was to determine if plant production of the selected mixtures could be simu- lated in the laboratory. All of the WMA mixtures produced in this study used the companion HMA mix designs, com- monly referred to as the “drop-in approach” for WMA, as is the common practice in most states. Therefore, only the HMA mixtures were verified in this study. Larsen, Wisconsin Table 5-1a and Table 5-1b show the JMF, the average mea- sured field gradations obtained by NCAT personnel, and grada- tion checks of laboratory-batched samples (verification). The mix design verification targeted the as-produced gradation measured by NCAT rather than the JMF gradation to match the produced mixture as closely as possible. For this project, the design asphalt contents were selected at 4.0% air voids at 75 Ndesign gyrations. Despite the use of cold-feed percentages to produce the aggregate blend, the percent passing individ- ual sieve sizes were not significantly different (less than 3%). Table 5-2 shows the aggregate specific gravities of individ- ual components and the methods used to determine these parameters. Several methods were used to obtain the specific gravity of RAP and RAS samples. The panel had suggested adding 1.5% and 15% extra asphalt binder to the RAS prior to conducting AASHTO T 209 to mitigate the amount of floating particles present during the AASHTO T 209 Test. However, it was found that the samples with 15% added binder were very difficult to handle and the results were not sensible, so the 15% was lowered to 5% for testing of RAS from the last three proj- ects. The combined aggregate specific gravity was calculated for the blend by using the RAS Gse, based on AASHTO T 209 without extra asphalt binder and the RAP Gsb from ignition- recovered aggregates. This combination process was used on all mixes verified in this study. Table 5-3 shows the volumetric properties at four asphalt contents, while Table 5-4 shows the volumetric properties at the optimum asphalt content. JMF- and NCAT-measured volumetric properties are also shown for comparison. Overall, the HMA design verification yielded a 0.6% higher asphalt content than the contractor’s mix design. However, the adjusted minimum VMA criterion in PP 78-17 was signifi- cantly exceeded. The higher VMA results for the mix design verification were largely caused by the higher combined aggre- gate Gsb results, compared to the combined Gsb value reported by the contractor. In addition, the parameter DTc determined in accordance with AASHTO PP 78-17 was −6.5, slightly below the minimum required value of −5.0. Enterprise, Alabama Table 5-5a and Table 5-5b show the JMF gradation, average QC values, and gradation checks of laboratory-batched sam- ples for the two Alabama HMA mixtures (adjusted air void C H A P T E R 5

125 Sieve Size Percent Passing JMF NCAT QC Verification 1 in. 100.0 100.0 100.0 ¾ in. 100.0 100.0 100.0 96.9 96.8 95.9½ in. 89.3 88.7 89.4 No. 4 76.2 74.7 76.3 No. 8 55.9 55.0 58.3 No. 16 41.7 41.0 43.8 No. 30 30.9 30.9 32.4 No. 50 16.1 16.3 16.9 No. 100 6.8 6.9 7.0 No. 200 4.2 4.3 4.2 in.83⁄ Table 5-1a. Wisconsin design, field, and verification gradations. Material Cold-Feed Percentages JMF NCAT QC Verification -in.5⁄8 x ½-in. chips 11.0 10.0 10.0 ½-in. x ¼-in.chips 8.0 9.0 9.0 Manufactured sand 36.0 35.0 35.0 Natural sand 27.5 29.0 28.5 Wisconsin baghouse fines 0.5 0.0 0.5 Wisconsin RAP 14.0 14.0 14.0 Wisconsin RAS 3.0 3.0 3.0 Table 5-1b. Wisconsin design material and cold-feed percentages. Material Test Method Gsb Gsa Absorption (%) ½-in. x ¼-in. chips AASHTO T 85 2.724 2.809 1.10 -in. x ½-in. chips AASHTO T 85 2.746 2.808 0.80 Combined JMF Multiple 2.699 2.767 0.65 Manufactured sand AASHTO T 84 2.695 2.803 1.45 Natural sand AASHTO T 84 2.668 2.689 0.30 RAP aggregate AASHTO T 84 ignition 2.682 2.759 1.05 AASHTO T 84 TCE extract 2.738 2.813 1.00 RAS aggregate ASTM D6857 CoreLok 2.581a AASHTO T 209 with 1.5% extra asphalt content 2.740a AASHTO T 209 with 15% extra asphalt content 2.856a AASHTO T 209 without extra asphalt content 2.633a aGse calculated from Gmm. 5⁄8 Table 5-2. Wisconsin aggregate specific gravities of individual components. Asphalt Content (%) Gmm Gmb Air Void (%) VMA (%) VFA (%) D/B 4.8 2.550 2.347 8.0 17.0 53.1 1.06 5.3 2.531 2.377 6.1 16.4 62.9 0.94 5.8 2.511 2.400 4.4 16.0 72.4 0.85 6.3 2.492 2.418 3.0 15.8 81.3 0.77 Optimum (6.0) 2.505 2.405 4.0 15.9 75.0 0.83 Table 5-3. Summary of NCAT design verification for Wisconsin.

126 Material Cold-Feed Percentages JMF Adjusted Air Void Low Air Void NCAT QC Verification NCAT QC Verification ½-in. crushed gravel 39.0 36.0 36.0 36.0 36.0 Shot gravel 11.0 8.0 8.0 8.0 8.0 78 limestone 7.0 10.0 10.0 10.0 10.0 Limestone screens 7.0 7.0 6.0 7.0 6.0 Sand 15.0 19.0 19.0 19.0 19.0 Baghouse fines 1.0 0.0 1.0 0.0 1.0 RAP 15.0 15.0 15.0 15.0 15.0 RAS 5.0 5.0 5.0 5.0 5.0 Table 5-5b. Alabama design material and cold-feed percentages. Property JMF NCAT QC Verification AASHTO M 323 Criteria AASHTO PP 78-17 Criteria Va (%) 4.0 4.4 4.0 Pb (%) 5.4 5.5 6.0 VMA (%) 14.5 15.5 15.9 Minimum 14.0 Minimum 14.3 VFA (%) 72.4 71.7 75.0 65–75 D/B ratio 0.90 0.90 0.83 0.6–1.2 RAP–BR 0.12 NA 0.15 RAS–BR 0.11 NA 0.13 Gmm 2.510 2.515 2.505 Gmb 2.410 2.404 2.405 Gsa NL 2.758 2.758 Gse 2.732 2.741 2.756 Gsb 2.668 2.691 2.691 Pba 0.91 0.70 0.90 Pbe 4.54 4.79 5.11 Tc (°C) NL -3.5 a -6.5 Minimum -5.0 Note: BR = binder ratio; NA = not available; NL = not listed. a20-h PAV. Table 5-4. Summary of Wisconsin volumetric properties. Sieve Size JMF Percent Passing Adjusted Air Void Low Air Void NCAT QC Verification NCAT QC Verification 1 in. 100.0 100.0 100.0 100.0 100.0 100.0 99.8 99.7 100.0 99.7 97.0 94.6 95.0 94.1 95.0 90.0 85.2 85.8 84.8 85.8 No. 4 64.0 61.6 61.2 60.6 61.2 No. 8 48.0 44.0 46.0 43.8 46.0 No. 16 39.0 33.9 37.6 34.1 37.6 No. 30 29.0 25.3 28.3 25.2 28.3 No. 50 16.0 14.3 16.4 14.4 16.4 No. 100 10.0 7.5 9.6 7.9 9.6 No. 200 6.5 4.7 6.4 5.2 6.4 ¾ in. ½ in. in.83⁄ Table 5-5a. Alabama design, field, and verification gradations.

127 and low air void). For this project, the JMF design asphalt content was selected at 4.1% air voids at 60 Ndesign gyrations. The second mixture used the same JMF but was intended to be produced at a target air voids content of 3.0%. Table 5-6 shows aggregate specific gravities of individual components and the procedures used to determine these parameters. Table 5-7 shows the volumetric properties at three asphalt contents, while Table 5-8 shows the volumetric properties at the optimum asphalt content. The JMF and NCAT quality control volumetric properties are also shown for comparison purposes. Overall, the mix design verification yielded similar asphalt content as the contractor’s mix design. However, the DTc results determined in accordance with AASHTO PP 78-17 were −13.8 and −10.6 for the adjusted and low air void mixtures, respectively, failing to meet the minimum required value of −5.0. In addition, the mix designed to a target of 4.1% air voids had a VMA of 14.0%, which is the minimum VMA criterion for AASHTO M 343 but does not satisfy the adjusted minimum VMA required in PP 78-17. The mix designed to a target of 3.0% air voids also failed to meet both VMA criteria; however, it does have 10.8% Vbe (VMA-Va = 13.8%-3.0%) which would meet the intent of PP 78-17 of having at least 10.5% Vbe. Oak Ridge, Tennessee Table 5-9a and Table 5-9b show the JMF, measured field gradations, and gradation checks of laboratory-batched sam- ples for the mix design verification. For this project, the design asphalt content was selected at 3.8% air voids at 75 blows with the Marshall hammer. Table 5-10 shows aggregate spe- cific gravities of individual components and the methods used to determine these parameters. Table 5-11 shows the volumetric properties at three asphalt contents, and Table 5-12 shows the volumetric properties at the optimum asphalt content. The field volumetric properties are also shown for comparison. The mix design verification for this project yielded a 0.3% higher optimum asphalt con- tent than the contractor’s mix design. The adjusted minimum VMA criterion of 14.3% was exceeded by a large margin; Material Test Method Gsb Gsa Absorption (%) ½-in. Crushed gravel AASHTO T 84 2.633 2.644 0.15 AASHTO T 85 2.608 2.652 0.60 Weighted average 2.615 2.649 0.30 78 limestone AASHTO T 85 2.729 2.758 0.40 Limestone screenings AASHTO T 84 2.667 2.747 1.10 Sand AASHTO T 84 2.618 2.642 0.35 Shot gravel AASHTO T 85 2.610 2.653 0.65 Combined JMF Multiple 2.640 2.671 0.38 RAP AASHTO T 84 TCE extract 2.651 2.682 0.40 AASHTO T 85 TCE extract 2.661 2.692 0.45 Weighted average 2.651 2.686 0.40 RAS ASTM D6857 CoreLok 2.708a AASHTO T 209 with 1.5% extra asphalt content 2.735a AASHTO T 209 with 15% extra asphalt content 2.964a AASHTO T 209 without extra asphalt content 2.733a aGse calculated from Gmm. Table 5-6. Alabama aggregate specific gravities of individual components. Asphalt Content (%) Gmm Gmb Va (%) VMA (%) VFA (%) D/B 4.5 2.486 2.375 4.5 14.1 68.3 1.53 5.0 2.468 2.395 3.0 13.8 78.6 1.37 5.5 2.450 2.404 1.9 14.0 86.6 1.24 Adjusted Va optimum (4.6%) 2.482 2.380 4.1 14.0 70.8 1.50 Low Va optimum (5.0%) 2.466 2.392 3.0 13.8 78.6 1.37 Table 5-7. Summary of NCAT design verification for Alabama.

128 Property JMF Adj. Va Low Va NCAT QC Verification NCAT QC Verification AASHTO M 323 Criteria AASHTO PP 78-17 Criteria Va (%) 4.1 3.0 4.1 1.8 3.0 Pb (%) 5.1 4.8 4.6 5.1 5.0 VMA (%) 15.5 13.3 14.0 12.9 13.8 Minimum 14.0 Minimum 14.5 VFA (%) 73.5 77.2 70.8 86.2 78.6 65–75 D/B ratio 1.29 1.05 1.50 1.10 1.37 0.6–1.2 RAP–BR 0.14 NA 0.15 NA 0.14 RAS–BR 0.20 NA 0.21 NA 0.19 Gmm 2.459 2.480 2.482 2.467 2.466 Gmb 2.358 2.405 2.380 2.423 2.392 Gsa NA 2.671 2.671 2.671 2.671 Gse 2.654 2.666 2.664 2.663 2.664 Gsb 2.650 2.640 2.640 2.640 2.640 Pba 0.05 0.39 0.35 0.34 0.35 Pbe 5.04 4.43 4.28 4.77 4.71 Tc (°C) NL -10.8 a -13.8 -7.7a -10.6 Minimum -5.0 Note: NA = not available; NL = not listed. a20-h PAV. Table 5-8. Summary of Alabama volumetric properties. Sieve Size Percent Passing JMF NCAT QC Verification 1 in. 100.0 100.0 100.0 100.0 100.0 100.0 98.0 96.9 96.9 88.0 83.7 83.7 No. 4 59.0 52.4 55.3 No. 8 41.0 38.2 39.8 No. 16 NA 29.7 29.9 No. 30 23.0 20.8 19.6 No. 50 14.0 12.7 11.6 No. 100 7.9 7.7 7.5 No. 200 5.1 5.2 5.5 Note: NA = not available. ¾ in. ½ in. in.83⁄ Table 5-9a. Tennessee design, field, and verification gradations. Material Cold-Feed Percentages JMF NCAT QC Verification D-Rock 42.0 42.3 42.3 Coarse slag 10.0 10.6 10.6 Limestone No. 10s 10.0 9.0 9.0 Natural sand 25.0 24.4 24.4 RAP 10.0 10.6 10.6 RAS 3.0 3.1 3.1 Table 5-9b. Tennessee design material and cold-feed percentages.

129 Material Test Method Gsb Gsa Absorption (%) Natural sand AASHTO T 84 2.663 2.738 1.02 D-Rock AASHTO T 85 2.736 2.783 0.61 Coarse slag AASHTO T 84 3.189 3.645 3.92 AASHTO T 85 3.257 3.476 1.94 Weighted average 3.226 2.551 2.51 Limestone No. 10s AASHTO T 84 2.751 2.840 1.15 Combined JMF Multiple 2.754 2.836 1.02 RAP aggregate AASHTO T 84 TCE extract 2.645 2.747 RAS aggregate ASTM D6857 CoreLok 2.069 2.069 AASHTO T 209 with 1.5% extra asphalt content 2.714a AASHTO T 209 without extra asphalt content 2.746a AASHTO T 209 with 5% extra asphalt content 2.733a aGse calculated from Gmm. Table 5-10. Tennessee aggregate specific gravities of individual components. Property JMF NCAT QCa Verification AASHTO M 323 Criteria AASHTO PP 78-17 Criteria Va (%) 3.8 4.4 3.8 Pb (%) 5.7 5.5 6.0 VMA (%) 17.4 14.4 16.1 Minimum 14.0 Minimum 14.3 VFA (%) 78.3 69.6 76.6 65–75 D/B ratio 0.90 1.26 1.08 0.6–1.2 RAP–BR 0.11 NA 0.11 RAS–BR 0.10 NA 0.09 Gmm 2.543 2.596 2.557 Gmb 2.447 2.482 2.460 Gsa 2.841 2.836 2.836 Gse 2.792 2.826 2.829 Gsb 2.752 2.754 2.754 Pba 0.53 0.95 0.99 Pbe 5.20 4.15 5.12 Tc (°C) NL -11.7 b -10.4 Minimum -5.0 Note: NA = not available; NL = not listed. aBased on the Marshall Method. b20-h PAV. Table 5-12. Summary of Tennessee volumetric properties. Asphalt Content (%) Gmm Gmb Va (%) VMA (%) VFA (%) D/B 4.8 2.608 2.373 9.0 18.0 49.8 1.42 5.3 2.588 2.402 7.2 17.4 58.7 1.26 5.8 2.567 2.439 5.0 16.5 70.0 1.13 6.3 2.547 2.481 2.6 15.6 83.4 1.02 Optimum (6.0) 2.557 2.460 3.8 16.1 76.6 1.08 Table 5-11. Summary of NCAT design verification for Tennessee.

130 however, its VFA was above the allowable range of AASHTO M 323. Although NCAT’s Gsb results were similar to those of the contractor, NCAT’s Gse results were much higher, indicat- ing a much higher asphalt absorption for the NCAT results. The DTc obtained in accordance with AASHTO PP 78-17 was −10.4, well below the minimum required value of −5.0. This mixture had a RAS binder ratio of 0.09. Based on Note 13 in PP 78-17, agencies may set a maximum RAS–BR at which no DTc testing is required. The recommended default minimum RAS–BR for this exception is 0.10. Wilson, North Carolina Table 5-13a and Table 5-13b show the JMF, measured field gradations, and gradation checks of laboratory-batched samples for the two North Carolina mixtures (MW–RAS and PC–RAS mixtures). The laboratory verification of the HMA mixture targeted the as-produced gradations rather than the JMF gradations. For this project, the optimum asphalt content was selected at 4.0% air voids and 65 Ndesign gyrations. Table 5-14 shows the aggregate specific gravities of indi- vidual components. Table 5-15 and Table 5-16 show the volu - metric properties at three asphalt contents for the mix design verifications for the MW–RAS mixture and the PC–RAS mix- ture, respectively. Table 5-17 shows the volumetric properties at the verified optimum asphalt content. The field volumetric properties are also shown for comparison. The mix design verification indicated a 0.2% higher optimum asphalt content for the MW–RAS mixture and 0.4% higher for the PC–RAS mix- ture compared to the contractor’s mix designs. The adjusted minimum VMA criterion of 15.5% was just barely met for the MW–RAS mixture but easily satisfied for the PC–RAS mixture. The DTc results were −10.5 and −7.3 for the verified MW–RAS and PC–RAS mixtures, respectively. Both failed the minimum required value of −5.0. La Porte, Indiana Table 5-18a and Table 5-18b show the JMF, measured field gradations, and gradation checks of laboratory-batched samples. For this project, the optimum asphalt content were selected at 4.0% air voids at 75 Ndesign gyrations. Table 5-19 shows the aggregate specific gravities of individual compo- nents and the methods used to determine these parameters. Sieve Size Percent Passing MW–RAS PC–RAS JMF NCAT QC Verification JMF NCAT QC Verification 1 in. 100.0 100.0 100.0 100.0 100.0 100.0 ¾ in. 100.0 100.0 100.0 100.0 100.0 100.0 ½ in. 100.0 99.1 99.7 100.0 98.8 99.7 96.0 93.1 94.8 96.0 92.5 94.8 No. 4 72.0 69.7 74.0 72.0 69.3 73.9 No. 8 57.0 54.3 60.0 57.0 53.7 59.8 No. 16 42.0 41.8 47.5 42.0 41.7 47.5 No. 30 29.0 28.9 33.6 29.0 29.0 33.7 No. 50 16.0 15.5 18.8 17.0 16.6 19.0 No. 100 10.0 7.7 10.3 10.0 8.5 10.5 No. 200 6.2 4.6 6.8 6.2 5.3 6.9 3⁄8 in. Table 5-13a. North Carolina design, field, and verification gradations. Material Cold-Feed Percentages MW–RAS PC–RAS JMF NCAT QC Verification JMF NCAT QC Verification 78-M stone 29 26.25 26.25 29 26.25 26.25 Dry screenings 13 17.75 17.75 19 17.75 17.75 Coarse sand 32 30 30 26 30 30 Baghouse fines 1 1 1 1 1 1 RAP 20 20 20 20 20 20 RAS 5 5 5 5 5 5 Table 5-13b. North Carolina design material and cold-feed percentages.

131 Material Test Method Gsb Gsa Absorption (%) RAP AASHTO T 84 2.616 2.658 0.60 Sand AASHTO T 84 2.610 2.656 0.66 No. 78s granite AASHTO T 85 2.608 2.649 0.59 Screenings AASHTO T 84 2.643 2.657 0.20 Combined JMF–MW–RAS Multiple 2.620 2.655 0.51 Combined JMF–PC–RAS Multiple 2.622 2.657 0.51 MW–RAS aggregate ASTM D6857 CoreLok 2.700a AASHTO T 209 with 1.5% extra asphalt content 2.690a AASHTO T 209 without extra asphalt content 2.671a AASHTO T 209 with 5% extra asphalt content 2.719a PC–RAS aggregate ASTM D6857 CoreLok 2.759a AASHTO T 209 with 1.5% extra asphalt content 2.756a AASHTO T 209 without extra asphalt content 2.714a AASHTO T 209 with 5% extra asphalt content 2.823a aGse calculated from Gmm. Table 5-14. North Carolina aggregate specific gravities of individual components. Asphalt Content (%) Gmm Gmb Va (%) VMA (%) VFA (%) D/B 4.9 2.464 2.309 6.3 16.2 61.1 1.53 5.4 2.447 2.337 4.5 15.6 71.2 1.38 5.9 2.429 2.355 3.0 15.4 80.2 1.25 Optimum (5.6) 2.440 2.343 4.0 15.5 74.5 1.33 Table 5-15. Summary of NCAT design verification for North Carolina MW–RAS. Asphalt Content (%) Gmm Gmb Va (%) VMA (%) VFA (%) D/B 4.9 2.465 2.273 7.8 17.5 55.7 1.56 5.4 2.446 2.317 5.3 16.4 67.8 1.40 5.9 2.429 2.338 3.8 16.1 76.7 1.27 Optimum (5.8) 2.432 2.335 4.0 16.1 75.1 1.30 Table 5-16. Summary of NCAT design verification for North Carolina PC–RAS.

132 Property MW–RAS PC–RAS JMF NCAT QC Verification JMF NCAT QC Verification AASHTO M 323 Criteria AASHTO PP 78-17 Criteria Va (%) 4.0 6.4 4.0 4.0 4.2 4.0 Pb (%) 5.4 5.0 5.6 5.4 5.4 5.8 VMA (%) 16.0 16.5 15.5 16.1 15.8 16.1 Minimum15.0 Minimum 15.5 VFA (%) 75.0 61.3 74.5 74.9 73.2 75.1 65–75 D/B ratio 1.16 1.00 1.33 1.16 1.04 1.30 0.6–1.2 RAP–BR 0.18 NA 0.21 0.18 NA 0.20 RAS–BR 0.17 NA 0.16 0.19 NA 0.14 Gmm 2.448 2.459 2.440 2.447 2.436 2.432 Gmb 2.350 2.301 2.343 2.349 2.333 2.335 Gsa 2.677 2.655 2.655 2.676 2.657 2.657 Gse 2.653 2.649 2.652 2.652 2.637 2.656 Gsb 2.648 2.620 2.620 2.647 2.622 2.622 Pba 0.10 0.44 0.48 0.10 0.22 0.50 Pbe 5.32 4.58 5.13 5.33 5.15 5.33 Tc (°C) NL -2.7 a -10.5 NL -3.2a -7.3 Minimum -5.0 Note: NA = not available; NL = not listed. a20-h PAV. Table 5-17. Summary of North Carolina volumetric properties. Sieve Size Percent Passing JMF NCAT QC Verification 1 in. 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 99.9 93.0 93.8 94.1 No. 4 67.1 68.7 68.7 No. 8 43.6 46.3 44.9 No. 16 27.5 30.6 28.5 No. 30 18.4 20.6 17.5 No. 50 12.1 12.6 10.8 No. 100 7.9 8.6 7.4 No. 200 5.9 6.7 5.6 ¾ in. ½ in. in.83⁄ Table 5-18a. Indiana design, field, and verification gradations. Material Cold-Feed Percentages JMF NCAT QC Verification Limestone No.11 28.0 25.0 25.0 No. 11 slag 7.0 13.0 13.0 Fine mineral 20 coarse 25.0 35.0 35.0 QA 24 slag 17.0 4.0 4.0 23/24 natural sand 6.0 6.0 6.0 Indiana RAP 15.0 15.0 15.0 Indiana RAS 2.0 2.0 2.0 Table 5-18b. Indiana design material and cold-feed percentages. Table 5-20 shows the volumetric properties at three asphalt contents. Table 5-21 shows the volumetric properties at the asphalt content used to bracket the field-measured asphalt con- tent. The field volumetric properties are also shown for com- parison. The mix design verification indicated a 0.8% higher optimum asphalt content than the contractor’s mix design. The adjusted minimum VMA criterion of 15.2% was met. However, the DTc of this mixture was −9.5, failing the minimum required value of −5.0. This mixture had a RAS binder ratio of 0.06. Based on Note 13 in PP 78-17, agencies may set a maximum RAS–BR at which no DTc testing is required. The recommended default minimum RAS–BR for this exception is 0.10. Property Comparisons Figure 5-1 shows asphalt content obtained from the JMF, QC results, and mix design verifications. Overall, higher asphalt contents were obtained from the mix design verifications, except for one of the Alabama mixtures. With regard to the effective volume of asphalt binder, Figure 5-2 shows that for most projects, Vbe values were lower for the mix design verifi- cation results compared to the respective JMF. On the other hand, production variability may have been responsible for the significant decrease in Vbe on the Tennessee and the North Carolina MW–RAS plant-produced mix samples. Table 5-22 shows DTc results for the recovered binders from the plant mix samples and laboratory-prepared mix design verification samples. Binders recovered from the plant mix

133 Material Test Method Gsb Gsa Absorption (%) Limestone No. 11 AASHTO T 85 2.772 2.820 0.61 No. 11 slag AASHTO T 85 2.424 2.615 3.03 Fine mineral 20s coarse AASHTO T 84 2.672 2.824 2.02 Slag sand AASHTO T 84 2.758 2.894 1.71 Combined JMF Multiple 2.657 2.777 1.64 Natural sand AASHTO T 84 2.621 2.744 1.71 RAP aggregate AASHTO T 84 2.652 2.756 1.42 RAS aggregate ASTM D6857 CoreLok 2.633a AASHTO T 209 with 1.5% extra asphalt content 2.638a AASHTO T 209 without extra asphalt content 2.636a AASHTO T 209 with 5% extra asphalt content 2.675a aGse calculated from Gmm. Table 5-19. Indiana aggregate specific gravities of individual components. Asphalt Content (%) Gmm Gmb Va (%) VMA (%) VFA (%) D/B 5.8 2.495 2.320 7.0 17.7 60.4 1.18 6.3 2.476 2.384 3.7 15.9 76.7 1.06 6.8 2.458 2.424 1.4 15.0 90.7 0.97 Optimum (6.3) 2.476 2.377 4.0 15.9 76.0 1.07 Table 5-20. Summary of NCAT design verification for Indiana. Property JMF NCAT QC Verification AASHTO M 323 Criteria AASHTO PP 78-17 Criteria Va (%) 4.0 5.3 4.0 Pb (%) 6.0 5.5 6.3 VMA (%) 15.6 16.0 15.9 Minimum 15.0 Minimum 15.2 VFA (%) 74.4 67.1 76.0 65–75 D/B ratio 1.30 1.44 1.07 0.6–1.2 RAP–BR 0.13 NA 0.13 RAS–BR 0.06 NA 0.06 Gmm 2.500 2.494 2.476 Gmb 2.401 2.363 2.377 Gsa NL 2.777 2.777 Gse 2.752 2.720 2.735 Gsb 2.675 2.657 2.657 Pba 1.06 0.90 1.11 Pbe 5.01 4.68 5.26 Tc (°C) NL -5.6 a -9.5 Minimum -5.0 Note: BR = binder ratio; NA = not available; NL = not listed. a20-h PAV. Table 5-21. Summary of Indiana volumetric properties.

134 Laboratory Verification NC MW–RASAL Adj. Va AL Low Va NC PC–RAS A sp ha lt Co nt en t ( % ) N C M W –RA S N C PC–RA S A sp ha lt Co nt en t D iff er en ce (% ) JMF–Laboratory Verification QC–Laboratory Verification A L A dj. V a A L Low V a Figure 5-1. Asphalt content comparisons. samples were subjected to 20-h PAV before testing, whereas 40-h PAV was used for the laboratory mix design verification samples. The change to 40-h PAV conditioning was made to AASHTO PP 78-17, based primarily on a study by Reinke et al. (2016) and other preliminary unpublished studies (Aschenbrener 2016). All of the recovered binders subjected to 40-h PAV failed to meet the minimum −5°C DTc limit in PP 78-17; even the mixtures with only 2% to 3% RAS. All of these mixtures also contained RAP, which contributed to the low DTc values. However, given the magnitude of the results (even if the mixtures only contained RAS), it seems unlikely that many would pass the current recommended minimum DTc. Therefore, mix designers would have to explore other means to increase the DTc values, such as using a softer grade of virgin binder or possibly adding a rejuvenator. Except for the Tennessee mixture, the DTc results for 40-h PAV aged binders from the laboratory mixtures were lower than the corresponding binders from plant-produced NC MW–RAS NC PC–RAS N C M W –RA S N C PC–RA S JMF–Laboratory Verification QC–Laboratory Verification Laboratory VerificationQCJMF V b e (% ) V b e D iff er en ce (% ) A L A dj. V a A L Low V a AL Adj. Va AL Low Va Figure 5-2. Effective volume of binder comparisons.

135 Location RAS Percentage and Type Binder Ratios Tc RAP RAS 20-h PAV 40-h PAV Wisconsin 3% PC 0.15 0.13 -3.5 -6.5 Alabama, Low Va 5% PC 0.14 0.19 -7.7 -10.6 Alabama, Adj. Va 5% PC 0.15 0.21 -10.8 -13.8 Tennessee 3% PC 0.11 0.09 -11.7 -10.4 North Carolina MW–RAS 5% MW 0.21 0.16 -2.7 -10.5 North Carolina PC–RAS 5% PC 0.20 0.14 -3.2 -7.3 Indiana 2% MW 0.13 0.06 -5.6 -9.5 Table 5-22. DTc: 20-h PAV versus 40-h PAV. -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.05 WI AL Adj. Va AL Low Va TN NC MRAS NC PRAS IN G m m D iff er en ce JMF–Laboratory Verification QC–Laboratory Verification Figure 5-3. Differences in Gmm. mixtures after 20-h PAV. Recall that for the Tennessee project, the plant mix HMA was stored in a silo for 4 h because of a rain delay, which could explain the higher value obtained at the 20-h condition. In general, maximum specific gravity tends to be one of the most repeatable tests. However, Gmm results are affected by binder absorption and therefore can be sensitive to differ- ences in aging or curing between laboratories. Figure 5-3 shows the differences in Gmm between contractor JMFs and mix design verification samples and between field QC results and mix design verifications for all of the mixtures evaluated. All of the mix verification samples were aged for 2 h at the field compaction temperature. Absolute differences were less than 0.025 with no defined tendencies for JMF versus verified mixtures. Greater differences were obtained when comparing Gmm from QC and verified samples, which is part of the vari- ability expected during production. Pba is calculated using the aggregate bulk and effective spe- cific gravities. The effective specific gravity is calculated using the mixture’s maximum specific gravity and asphalt content. Therefore, differences in aggregate specific gravities or the mix- ture’s maximum specific gravity will affect the reported Pba. Figure 5-4 shows the differences in Pba between JMF and mix design verifications and between QC results and mix design verifications. In this case, mixture verification provided higher percent binder absorption than the JMF and field QC results. This could be explained by the higher Gse values obtained on the mixture design verification samples (Figure 5-5) and the higher asphalt binder content obtained with the mixture veri- fication for all projects, except for the Alabama adjusted air void mixture. Ideally, the laboratory design should be able to replicate the field-produced material with regard to volumetric properties. Differences in gradation can lead to differences in volumetric properties, and the JMF is not always reproduced in the field. However, some properties such as Gse and Gsb have significant effects on volumetric properties that are performance-related. Figure 5-6 exhibits differences in VMA where verified mix- tures had lower VMA values in most cases. Consequently, lower VFA values were obtained during verifications for all

136 P b a D iff er en ce (% ) JMF–Laboratory Verification QC–Laboratory Verification NC PC–RASNC MW–RASAL Adj. Va AL Low Va Figure 5-4. Differences in binder absorption. G se D iff er en ce (% ) JMF–Laboratory Verification QC–Laboratory Verification NC PC–RASNC MW–RASAL Adj. Va AL Low Va Figure 5-5. Differences in Gse. VM A D iff er en ce (% ) NC PC–RASNC MW–RAS JMF–Laboratory Verification QC–Laboratory Verification AL Adj. Va AL Low Va Figure 5-6. Differences in VMA.

137 mixtures. Once again, this could be an indication that mix- tures were potentially placed too dry, which may increase cracking susceptibility. Summary In general, the results of the mix design verifications indi- cated slightly higher asphalt content to meet the volumetric requirements for most of the mixtures in this study. In addi- tion, critical properties such as the effective specific gravity of the mixture and the specific gravity of the aggregate, which significantly affect volumetric properties (VMA and VFA), tended to have higher verified values. Determination of RAS aggregate Gsb can be difficult because of the small size of RAS aggregate particles. Based on the adjusted VMA criteria in AASHTO PP 78-17, only one of the six verified mixes had a failing VMA. This mixture would require changes in the combined aggregate blend and/or an increase in the binder content to meet the higher VMA requirement. None of the mixtures met the new DTc criterion of –5.0 in AASHTO PP 78-17. To meet this criterion, the mix design would have to be modified by reducing the amount of RAS or by using a softer virgin binder. More research is needed to validate the DTc criterion. VF A D iff er en ce (% ) NC PC–RASNC MW–RAS JMF–Laboratory Verification QC–Laboratory Verification AL Adj. Va AL Low Va Figure 5-7. Differences in VFA.