Click for next page ( 27


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 26
26 Table 21. Regression analyses between rutting parameters and coarse aggregate UVA. Predictor Response Variable Regression Equation R2 p-value Variable Total Rut Depth at UVA 164.54 3.19 UVA 0.96 0.004 5,000 Wheel Passes Early Rut Rate UVA 23.6277 0.44 UVA 0.95 0.005 Later Rut Rate UVA 90.14 1.78 UVA 0.96 0.004 grouped. The coarse aggregates with UVA values of 49 and 51 35 percent resulted in a rut depth of approximately 6 mm at UVA=42, N7=50 20,000 APT wheel passes; the two coarse aggregates with UVA 30 UVA=48, N7=858 values of 48 percent resulted in a total rut depth of about 10 mm after 20,000 passes; and the rounded gravel with a coarse UVA=48, N7=897 aggregate UVA of 42 percent resulted in a rut of 30 mm at 25 5,000 wheel passes. These data suggest that a minimum coarse UVA=49, N7=13,370 aggregate UVA value of 45 percent would be desirable. Rut Depth, mm 20 UVA=51, N7=6210 150 15 0.19x y = 0.0065e Number of APT Wheel Passes R2 = 0.45 to Reach 3.5 mm Total Rut 10 100 5 50 0 0 5000 10000 15000 20000 Number of Wheel Passes, N 0 40 45 50 55 Figure 21. Effect of coarse aggregate UVA on rut UVA, % depth. Figure 19. Effect of coarse aggre- gate UVA on traffic (3.5 mm). Fine-Graded Mixtures In NCHRP Report 405, correlations between fine aggregate properties and SST results were found to be poor; only Number of Wheel Passes to Reach 7 mm 20000 y = 10-9e0.58x the GLWT results were used for recommending fine aggregate R2 = 0.80 UVA as a predictor of mixture rutting performance. In the 15000 current study, HMA aggregates were tested using ASTM C 1252, Methods A and B, as well as VTM5. These three tests are Total Rut 10000 significantly correlated and give similar correlations with APT permanent deformation results. Rutting parameters used to evaluate fine-graded HMA 5000 mixtures' rutting potential were rut depths at 1,000; 5,000; and 20,000 APT wheel passes, and early and late traffic rut- 0 ting rates. Correlation analyses were conducted between the 40 45 50 55 rutting parameters and the fine aggregate test results. In addi- UVA, % tion to the UVA, UVB, and VTM5 tests, other aggregate tests Figure 20. Effect of coarse aggre- incorporated in the analysis were Micro-Deval (MDEV); gate UVA on traffic (7 mm). magnesium sulfate soundness (MGSO4); and D60, D30, and

OCR for page 26
27 Table 22. Correlation matrix between fine aggregate prop- erties and rutting parameters. UVA UVB VTM5 MDEV MGSO4 D60 D30 D10 1,000 -0.809 -0.847 -0.838 -0.393 -0.393 0.736 0.773 0.393 Total rut depth passes 0.051 0.033 0.037 0.441 0.441 0.096 0.071 0.441 5,000 -0.372 -0.511 -0.517 0.765 0.272 0.315 0.292 0.412 passes 0.538 0.379 0.372 0.132 0.659 0.606 0.634 0.490 20,000 -0.715 -0.666 -0.685 0.948 0.906 0.466 0.312 0.360 passes 0.286 0.334 0.315 0.052 0.094 0.534 0.688 0.641 Early Rut -0.793 -0.846 -0.839 -0.287 -0.354 0.727 0.762 0.326 Rate 0.0599 0.034 0.037 0.581 0.492 0.102 0.078 0.529 Later Rut -0.795 -0.822 -0.812 -0.483 -0.419 0.746 0.785 0.459 Rate 0.059 0.045 0.0499 0.332 0.409 0.088 0.064 0.360 Descriptive 6.4 7.0 7.0 5.0 4.2 5.6 5.2 3.4 Ranking D10 of the p0.075 fraction. The results are shown in Table 22. 20 Only UVA, UVB, and VTM5 produce consistent correlations with the rutting parameters, although the MDEV and MGSO4 show good correlations with rut depth at 20,000 15 @ 5000 Wheel Passes, mm wheel passes. Plots of the rutting parameters as a function of Total Rut Depth fine aggregate UVA are shown in Figures 22 through 26. Similar to the procedure used for the coarse-graded mix- 10 tures, each of the fine aggregate test methods was assigned a descriptive ranking as shown in Table 22. The ranking was determined in the manner described for the coarse-graded y = -0.73x + 44.44 mixtures. Table 22 shows that fine aggregate UVB and VTM5 5 R2 = 0.28 have the highest descriptive rankings at 7.0; the fine aggregate UVA is only slightly less at 6.4. The D60, D30, and MDEV rankings are the next closest at 5.6, 5.2, and 5.0, respectively. 0 Regression analyses were performed to develop equations 40 42 44 46 48 50 relating rutting parameters and aggregate test results. The pri- UVA, % mary independent variables were UVA, UVB, and VTM5. The Figure 23. Rut depth/fine aggregate response variables were rut depths at 1,000; 5,000; and 20,000 UVA relationship (5,000 passes). APT wheel passes; and early and late traffic rutting rates. 35 35 y = -2.08x + 106.21 y = -1.03x + 60.32 30 2 30 R2 = 0.46 R = 0.62 @ 20000 Wheel Passes, mm @ 1000 Wheel Passes, mm 25 25 Total Rut Depth Total Rut Depth 20 20 15 15 10 10 5 5 0 0 40 42 44 46 48 50 40 42 44 46 48 50 UVA, % UVA, % Figure 22. Rut depth/fine aggregate Figure 24. Rut depth/fine aggregate UVA relationship (1,000 passes). UVA relationship (20,000 passes).

OCR for page 26
28 10 Table 23. Regression analyses between rutting y = -0.41x + 22.38 parameters and fine aggregate UVA. R2 = 0.62 8 Response Predictor Model Equation R2 p-value Variable Variable Rut UVA 100.26 1.95 UVA 0.65 0.051 Depth @ Early Rut Rate 6 UVB 112.98 2.05 UVB 0.72 0.033 mm / log(N) 1,000 passes1 VTM5 108.88 1.92 VTM5 0.70 0.037 Rut UVA 34.52 0.52 UVA 0.14 0.538 4 Depth @ 5,000 UVB 47.88 0.74 UVB 0.26 0.379 passes2 VTM5 46.34 0.69 VTM5 0.27 0.372 2 Rut UVA 63.67 1.11 UVA 0.51 0.286 depth @ 20,000 UVB 66.29 1.06 UVB 0.44 0.334 passes3 VTM5 64.84 1.01 VTM5 0.47 0.315 0 40 42 44 46 48 50 UVA 22.29 0.42 UVA 0.63 0.060 Early UVA, % Rut UVB 25.40 0.45 UVB 0.72 0.034 Rate1 Figure 25. Early rut rate/fine aggre- VTM5 24.56 0.42 VTM5 0.70 0.037 gate UVA relationship. UVA 70.66 1.41 UVA 0.63 0.059 Later Rut UVB 78.86 1.46 UVB 0.68 0.045 25 Rate1 VTM5 75.82 1.36 VTM5 0.66 0.0499 y = -1.49x + 74.16 1Allmixtures are included R2 = 0.58 2FA-1 20 (Natural Sand A) is not included 3FA-1 (Natural Sand A) and FA-2 (Crushed Gravel Sand) are not included Late Rut Rate 15 mm / log(N) ance of an HMA mixture also increases; however, UVA may be preferable to UVB or VTM5 because the UVA test requires less 10 material and time. Total rut depths at 5,000 and 20,000 wheel passes have poor correlations with fine aggregate UVA, UVB, and VTM5 (see 5 Table 23). Although the correlations are poor, the plots indi- cate the effect of fine aggregate UVA, UVB, and VTM5 on rut- 0 ting performance. Close examination suggests increasing fine 40 42 44 46 48 50 aggregate UVA from 42 to approximately 47 percent does not UVA, % appear to improve HMA rutting resistance; however, rutting Figure 26. Late rut rate/fine aggregate resistance increases significantly for fine aggregate UVA values UVA relationship. above 47 percent. Kandhal and Parker (1) reported similar results. Based on their GLWT data, mixture rutting suscepti- bility does not change much for fine aggregate UVA values of Results of the regression analyses are given in Table 23. The approximately 45 to 46 percent, but mixtures become less sus- relationships between total rut depth at 1,000 wheel passes and ceptible to rutting for fine aggregate UVA values above 46 per- early and late traffic rutting rates and UVA, UVB, or VTM5 are cent. Furthermore, at UVA values of around 39 to 40 percent, good/fair and significant. These are the three rutting parame- HMA mixtures tested in both the GLWT and APT exhibited ters that made use of all six fine aggregates. Not all six fine significant increases in rut depths. aggregates were used for rut depth analysis because no rut Based on results in this research and those reported by depth data were obtained for the FA-1 mixture at 5,000 and Kandhal and Parker, it appears that fine aggregate UVA is a 20,000 APT wheel passes and for the FA-2 mixture at 20,000 good predictor variable of HMA rutting performance. How- wheel passes. Additional fine aggregate properties such as ever, it should be recognized that in-place properties of HMA MDEV and MGSO4 were included in the regression analyses pavements, such as pavement density, also affect pavement using stepwise regression, but these additions failed to performance. Furthermore, it is recommended that natural improve the R2 values. These results show that as fine aggre- sands with fine aggregate UVA values below 40 percent be gate UVA, UVB, or VTM5 values increase, the rutting resist- avoided, unless they are combined with a higher UVA fine

OCR for page 26
29 aggregate and/or pass a performance test. Fine-graded HMA gate UVA and the number of APT wheel passes required to mixtures with fine aggregates having UVA values between 42 reach a 7.5- and 10-mm total rut depth, respectively. Both rela- and 47 percent tend to exhibit similar rutting performance. tionships appear to be exponential and significant. The expo- Mixtures with fine aggregate UVA values above 47 percent nential regression equation shown in Figure 27 indicates the appear to exhibit better rutting performance. However, expe- number of APT wheel passes required to reach a 7.5-mm total rience has shown that an HMA mixture with a high fine rut depth increases by approximately 950 for an increase in fine aggregate UVA can produce a mixture with high voids in the aggregate UVA from 45 to 50 percent. For this same 5-percent mineral aggregate (VMA). Such mixtures require a high increase in fine aggregate UVA, the equation shows the num- binder content to meet air voids criteria, which can lead to ber of APT wheel passes increases by approximately 6,600 to over-asphalting and poor mixture rutting performance. reach a 10-mm total rut depth. These relationships seem log- Sensitivity of the fine aggregate UVA to APT traffic levels ical up to a point. For example, for fine aggregate UVA values was analyzed at the 7.5-mm and 10-mm total rut depths. between 35 and 40 percent, the number of wheel passes Figures 27 and 28 show the relationships between fine aggre- applied before reaching a given rut depth increases very little. However, the increase is much larger for fine aggregate UVA 2500 values between 45 and 50 percent. The practical limit to the relationship is that many fine aggregates have UVA values of y = 8E-05e0.33x no more than approximately 50 percent. R2 = 0.57 2000 Figure 29 presents the APT rut depths as a function of the Number of APT Wheel Passes to Reach 7.5 mm Total Rut number of wheel passes for each of the fine aggregates used in the study. Unlike the coarse aggregate results, 1500 the fine aggregate UVA values are not grouped. Two of the fine aggregates have UVA values of 49 percent, but the gran- 1000 ite outperformed the traprock; the former yielding approx- imately a 5-mm rut depth at 20,000 wheel passes and the latter an 11-mm rut depth at 20,000 wheel passes. Natural 500 Sand B with a UVA of 42 percent performed as well as the dolomite sand with a UVA of 47 percent; both exhibited 0 40 42 44 46 48 50 35 UVA, % UVA=40, N10=75 Figure 27. Effect of fine aggregate UVA=46, N10=500 30 UVA on traffic (7.5 mm). UVA=42, N10=2500 UVA=49, N10>20,000 10000 25 UVA=47, N10=2500 y = 0.0002e0.35x UVA=49, N10=7500 9000 2 R = 0.51 Rut Depth, mm 8000 20 Number of APT Wheel Passes to Reach 10 mm Total Rut 7000 6000 15 5000 4000 10 3000 2000 5 1000 0 0 40 42 44 46 48 50 0 5000 10000 15000 20000 UVA, % Number of Wheel Passes, N Figure 28. Effect of fine aggregate Figure 29. Effect of fine aggregate UVA on rut UVA on traffic (10 mm). depth.