Annotated Literature Review for NCHRP Report 640 (2009)

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.

193 Table 99: OGFC Properties Material/property Type/value Binder PG 76-22 Asphalt binder content 5.5 Maximum size, mm 19 Nominal maximum aggregate size (NMAS), mm 12.5 Percent passing the 9.5 mm sieve 81 Percent passing the 4.75 mm sieve 14 Percent passing the 2.36 mm sieve 2 Percent passing the 1.18 mm sieve 1.4 Percent passing the 0.6 mm sieve 1.3 1.46.4 Construction Practices No information on construction practices has been presented. 1.46.5 Maintenance Practices No information is provided on maintenance practices of friction course. 1.46.6 Rehabilitation Practices No information is provided on rehabilitation practices of friction course. 1.46.7 Performance With respect to performance of OGFC, several observations were made from the data. Flintsch et al conducted macrotexture measurements on OGFC sections using CT Meter and laser profiler, conducted tests with laser profiler to estimate mean profile depths for validation of speed constant equation, used an existing equation to predict texture depth for unsegregated areas in HMA, and finally developed a new equation to predict mean texture depth in unsegregated areas in different types of HMA, including OGFC. In the comparison of results from different test methods, Flintsch et al showed that the macrotexture of the OGFC was the highest, as evident from results of sand patch tests, CT Meter and laser profiler tests. In their data for IFI speed constant equation validations, Flintsch et al showed that the OGFC sections had the highest macrotexture and the lowest percent normalized gradient (PNG). Note that the gradient is inversely proportional to the pavement macrotexture. 1.46.8 Structural Design No information is provided on structural design of friction course. 1.46.9 Limitations No information is provided on limitations of use. 1.47 Kaloush, K. E., M. W. Witczak, A. C. Sotil and G. B. Way. “Laboratory Evaluation of Asphalt Rubber Mixtures Using the Dynamic Modulus (E*) Test.” TRB 2003 Annual Meeting CD-ROM. Transportation Research Board. National Research Council. Washington, D.C. 2003.

194 1.47.1 General In this paper, Kaloush et al provides the results of a study conducted to evaluate the dynamic modulus of two asphalt rubber mixtures, one of which was an OGFC. The authors conducted testing at low and high temperatures, and developed master curves for these mixes. Kaloush et al compared the results for the OGFC with results obtained from tests on dense-graded mixes. They concluded that under unconfined conditions the asphalt rubber mixes showed lower modulus values and, hence, better resistance to cracking, whereas under confined conditions they showed higher modulus values and, hence, better resistance against rutting. However, the improvement in resistance at higher and lower temperatures was not as enhanced in the OGFC as it was in the other mix. On the other hand, the difference between modulus values obtained under confined and unconfined was most pronounced for the OGFC mixes. Kaloush et al concluded that the dynamic modulus test was appropriate for performance evaluation of different asphalt mixes, including asphalt rubber mixes (and OGFC mixes), but recommended that when comparing mixes with both dense and open gradations, the use of confined testing should be made. 1.47.2 Benefits of Permeable Asphalt Mixtures No information on benefits of permeable asphalt mixtures is provided. However, in the introduction Kaloush et al mention some of the benefits of using crumb rubber in different mixes, including OGFCs. These benefits include adequate resistance against reflection cracking, high temperature conditions of the Arizona desert, cold conditions at higher elevations (e.g. Flagstaff, AZ), reduction in tire noise, and improvement in ride quality. 1.47.3 Materials and Design Kaloush et al mention that samples of two different types of mixes were obtained during construction for laboratory testing. The mixes were Asphalt Rubber Asphalt Concrete (ARAC) Gap-Graded mixture, and Asphalt Rubber Asphalt Concrete Friction Course (AR-ACFC) Open-Graded mixture. The nominal aggregate size of the ARAC and the AR-ACFC mixes were 19.0 mm and 9.0 mm respectively, whereas the average air voids were 11 and 18 percent, respectively. The mixes were provided with a base binder grade of PG 58-22 grade. 1.47.4 Construction Practices Kaloush et al indicates that the samples of the two mixes were obtained from a construction project on I-40 (on the Buffalo Range paving project) east of Flagstaff, Arizona. They mention that these ARAC and AR-ACFC mixes, which are typically used in Northern Arizona, are placed in 50 mm and 12.5 mm thickness respectively. With the AR-ACFC mix placed on top of the ARAC mix. 1.47.5 Maintenance Practices No information is provided on maintenance practices of friction course.

195 1.47.6 Rehabilitation Practices No information is provided on rehabilitation practices of friction course. 1.47.7 Performance Kaloush et al provide background information on the dynamic modulus test procedure (ASTM D3497) that they used for determination of modulus, E* and phase angle, φ. Test specimens were obtained by coring 150mm tall by 100mm diameter samples from 150mm tall and 150mm diameter Superpave gyratory compacted samples. The test matrix is shown in Table 100. Table 100: Test Matrix Testing Condition/samples Values/numbers Confinement 0, 10, 20, and 30 psi Frequency 0.1, 0.5, 1, 5, 10, and 25 Hz Temperature 14, 40, 70, 100, and 130°F Number of Samples 6 for ARAC; 12 for ARACFC; 1 formed (molded) crumb rubber specimen consisting of 80% crumb rubber and 20% urethane. Sequence of Testing Each specimen was tested in an increasing order of temperature, i.e. 14, 40, 70, 100, and 130°F. For each temperature level, specimens were tested in a decreasing order of frequency (25, 10, 5, 1, 0.5, and 0.1 Hz). Note: The authors used the specimen instrumentation method that was developed by the Arizona State University Research Team. In this method, the LVDT’s were secured in place using brackets and studs glued on to the specimen; guiding rods were added to the instrumentation for alignment especially at high temperatures; the load was varied with temperature to keep the specimen response within a linear range (initial microstrains about 20-25 micro-strain). For the ARAC mix, Kaloush et al note that the comparison of the results for the unconfined and confined tests show that there is a significant increase in the E* values with confinement at higher temperatures and lower frequencies, compared to the low temperature part of the curve. The difference in E* results between the unconfined and confined tests at higher temperatures become less as the confinement is increased: 400percent increment from unconfined conditions to a 10-psi confinement, 25 percent from 10 to 20-psi confinement, and 11 percent increment from 20 to 30-psi confinement. They mention that the difference between the unconfined and confined tests at lower temperatures is much less, but still significant between the different levels of confinement. Kaloush et al indicate that there were cases where the E* value of the specimens at 100 or 130°F were equivalent. The authors hypothesize that this is due in large measure to the decreased role of the asphalt binder in relationship to the increased role of the rubber particles at higher temperatures. They attempted to prove this by conducting a series of tests on a crumb rubber specimen (see Table 100). From the E* results obtained from this specimen at 70, 100, 130°F, the authors note that the E* values

196 remain almost the same throughout the test at the different frequencies (loading time) and the three test temperatures. For the AR-ACFC mixes, Kaloush et al indicate that the trends of the results were similar to those of the ARAC mixes. They note a significant increase in the E* values with confinement at higher temperatures and lower frequencies, compared to the low temperature part of the curve. The authors mention that the difference in E* results between the unconfined and confined tests at higher temperatures stay significant as the confinement level is increased: 250 percent increment from unconfined conditions to a 10-psi confinement, 61 percent from 10 to 20-psi confinement, and 62 percent increment from 20 to 30-psi confinement. Kaloush et al point out an important property noted for the AR-ACFC mixes only - the difference between the different levels of confinements is negligible at lower (cold) temperatures, but the difference between the unconfined and confined tests is still significant (unlike the ARAC mix) at the lower temperatures. Kaloush et al provide the results of a study carried out to compare dynamic modulus values of asphalt rubber mixes (used in this study) versus dynamic modulus values of conventional ADOT dense graded mixes with PG 76-16 and PG 64-22 binders. The PG 76-16 mixture showed higher modulus values at all temperature and frequency conditions, whereas the E* values of the asphalt rubber mixes were found to be more comparable to those of the conventional PG 64-22 mixture. The authors note that these comparable values (asphalt rubber mixes showed higher modulus values at higher temperatures and lower modulas values at lower temperatures) were observed despite the fact that there was a significant air void difference between the asphalt rubber and the dense graded mixes. Kaloush et al concludes that this observation supports the superior performance of asphalt rubber mixes against high temperature rutting and low temperature cracking. They also note that the lower modulus values of the asphalt rubber mixes (at al temperatures) compare to the PG 76-16 mixes could very well be due to (at least partially) the impact of higher air voids in the asphalt rubber mixes. They recommend the comparison of results from samples with similar void levels. Kaloush et al then indicate that a comparison between different dynamic modulus values (dense-graded and asphalt rubber gap and open graded mixes) was conducted by selecting a reference mix. This comparison was carried out for results obtained at 14°F and 100°F, both at 10 Hz. The modular ratio (R) was calculated using the following equation R = E*mix/E*Reference, where R = Modular Ratio, E*mix = Dynamic Complex Modulus value for a given mixture, E*Reference = Dynamic Complex Modulus value for the reference mixture. For low temperature performance, the authors ranked the mixes in terms of increasing E* values (that is decreasing performance). For high temperature performance, the authors raked them in the reverse way – according to decreasing E * values (that is decreasing performance). First, the PG 64-22 conventional mix was taken as the reference mix. Kaloush et al mention that that asphalt rubber mixes have the lowest E* values (lowest modular ratio), at low temperatures. For the high temperature ranking, the asphalt rubber mixes were also found to have the lowest E* values – the authors point out that this may be due to unconfined testing and higher air