VTRC UPDATE continues on page 12 △ and 13% of failures, respectively. The failure of these variations indicates that the mixes were designed close enough to the mass loss criterion that aggregate gradation variability or lower asphalt content could cause the initial passing JMF to fail during production. The H-J variation provided the lowest values of mass loss 77% of the time. In addition, the H-J, H-C, and H-F variations improved the mass loss over the O-J mix in every case. This indicates that increased asphalt content is likely the most effective means of lowering mass loss. The impact of the gradation variations are less clear cut as their effects depend on the initial aggregate structure. CT Index Figure 2 shows the lab-produced Cracking Tolerance (CT) Index values overlaid with the production and reheat sample values. From design and production, one JMF and six reheat samples failed the CT index requirement of 70. The failing JMF (B 30R PG 64S-22) was a non-BMD mix. Note that these mixes predate the current production requirement of CT index greater than or equal to 95; however, if the production requirement of 95 were applied to these mixes, six production samples would fail the requirement, along with one production mix (A 40R PG 58-28). Except for one mix (E 35R PG 58-28 RA), average production CT index values were always greater than the average reheat CT index, likely due to the additional aging introduced during the reheating process. Average production values were also greater than JMF CT index values except for B 40R PG 64S-22 RA and C 35R PG 58-28. In seven of the thirteen cases, the reheat sample CT index values bracket the JMF value. For the remaining six cases, half had JMF CT index values lower than both production and reheat values, while for the remaining three, the production samples bracketed the JMF CT index value. Only three mixes (C 35R PG 58-28, E 35R PG 58-28 RA, and E 35R PG 58-28 Softening Oil + Fiber) had all lab-produced variations, production samples, and reheat samples pass the CT index criterion of 70. As with the mass loss, the lab-produced mixes show the influence of materials variability on the CT index results. Only 8 of the 13 O-J mixes passed the CT criterion of 70. Of those eight, four had CT index values less than the JMF while four had values greater than the JMF value. The L-J variation was most likely to fail the CT index criterion, with a 70% failure rate, while the O-C and O-F variations were the next most likely condition to fail, with each having a 46% failure rate. These results indicate that the optimum asphalt content is being selected too close to the CT index minimum value of 70 during design to provide assurance that a mix will not fail during production due to gradation or asphalt content varying within the acceptable tolerance limits. This is further indicated as the H-J variation improved the CT index values over that of the O-J variation for every mix. Additional analysis was performed to compare the average test result of each individual lab-produced mix variation (e.g., O-F) to the average O-J test result using the CT index difference two-sigma limit (d2s) precision estimates for single-operator and multi-laboratory conditions. This analysis was performed to investigate if the production tolerance limits for gradation and asphalt content produce test results that are wider or tighter than the variability of the CT index test method. Table 1 shows the results of this analysis. For most of the mixes, the test results between the O-J and variations in coarse and fine aggregates did not statistically differ, as seen in Table 1. Only 2 and 3 out of 14 observations were statistically significant for O-C and O-F variations, respectively. These results generally indicate that the test variability is similar or wider than the variation induced from the production tolerance limits on coarse and fine aggregates. This further indicates that the current gradation tolerance limits are appropriate to apply in the context of the IDT-CT test. Figure 2. CT Index. Dashed line indicates the minimum acceptable CT index. CT = cracking tolerance; R = RAP content; N/A = data not available; PG = performance grade; RA = recycling agent; JMF = job mix formula; O = optimum asphalt content; L = low asphalt content; H = high asphalt content; C = coarse gradation; F = fine gradation J = JMF gradation. Mix Variation Number (Percent) of Mixes Exceeding the d2s Boundaries Single Operator Multi-laboratory O-C 2 (14.3%) 0 (0.0%) O-F 3 (21.4%) 0 (0.0%) L-J 11 (78.6%) 1 (7.1%) L-C 1 (50.0%) 1 (50.0%) L-F 2 (100.0%) 1 (50.0%) H-J 9 (64.3%) 1 (7.1%) H-C 2 (100.0%) 0 (0.0%) H-F 0 (0.0%) 0 (0.0%) Table 1. Analysis Results Summary for Pairwise Comparison between O-J Mix and Its Variations. O = optimum asphalt content; L = low asphalt content; H = high asphalt content; C = coarse gradation; F = fine gradation; J = JMF gradation. VAASPHALT.ORG 11
RkJQdWJsaXNoZXIy Nzc3ODM=