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.
42 RLPD testing was conducted on cylindrical specimens with axially applied loading using an AMPT in general accordance with AASHTO TP 79. Modifications to the test included using a lower test temperature (45°C) and the same small-scale cylindrical specimen geometry as used for the dynamic modulus testing. A repeated deviator stress of 482.6 kPa was applied at a constant confining stress of 68.9 kPa. 5.1 RLPD Analysis Two methods of analysis, one qualitative and one quantitative, were considered for evaluating the RLPD data. The quantitative analysis regressed a power-law function fit to the RLPD curve between 2,000 and 10,000 cycles with the exception of one mixture and a few individual specimens that experienced premature termination. The slope and intercept values were calculated in a spreadsheet from the regressed power-law function for the outlier analysis used in the qualitative investigation. An example of this fit is shown in Figure 31. Upon removing outliers, RLPD envelopes were devel- oped to qualitatively examine the effects of recycling type, recycling agent, and chemical additives on the different in-place recycling mixes. A MATLAB code was used to extract slope and intercept values that were then used in the AASHTOWare Pavement ME Design® software for performance prediction. The performance prediction using RLPD values is discussed further in Chapter 6. 5.2 Outlier Analysis The slope values calculated in the spreadsheet are shown in Figure 32, and the intercept val- ues are shown in Figure 33. Each mixture type was grouped with its primary stabilizing additive and with its chemical additive if one was used. The outlier identification approach is similar to that described for determining dynamic modulus outlier values (see the section on outlier analysis in Chapter 4). Specimens were arranged by each mix type/stabilizing agent/chemical agent combination and the IQR was calculated for each grouping. If the dynamic modulus of a specimen was greater than the third quartile plus the IQR or less than the first quartile minus the IQR, that specimen was deemed an outlier. All specimens that fell between the outer limits were considered for further analysis. Of the 24 mixes used for analysis, one mix was removed before outlier analysis due to early catastrophic failure. Deficiencies in specimens, believed to be due to sampling and/or prepara- tion issues, led to the removal of five additional specimens (from three unique mixes), each of which never reached 2,000 cycles. The outlier analysis led to the removal of eight specimens over six mixes. Table 13 summarizes the number of mixes evaluated, specimens considered, and specimens remaining after outlier analysis. C H A P T E R 5 RLPD Test Results
RLPD Test Results 43 Figure 31. Example RLPD curve and power-law function fit with start and finish points indicated. BM = HMA base mixture, E = emulsiï¬ed asphalt, F = foamed asphalt, C = hydraulic cement, and L = lime. 0 0.2 0.4 0.6 0.8 1 1.2 CC PR E CC PR E- C CI R E CI R E- L CI R E- C CI R F CI R F- C FD R E FD R E - C FD R F -C BM Sl o pe Data Mean Quartile Figure 32. Slope of the secondary stage of the RLPD data and potential outliers.
44 Material Properties of Cold In-Place Recycled and Full-Depth Reclamation Asphalt Concrete BM = HMA base mixture, E = emulsiï¬ed asphalt, F = foamed asphalt, C = hydraulic cement, and L = lime. 0 2000 4000 6000 8000 10000 12000 CC PR E CC PR E- C CI R E CI R E- L CI R E- C CI R F CI R F- C FD R E FD R E - C FD R F -C BM In te rc ep t Data Mean Quartile Figure 33. Intercept of the secondary stage of the RLPD data and potential outliers. Mixture Type No. Mixtures No. Specimens Samples after Outlier Removal CCPR E 1 3 3 CCPR E-C 2 6 4 CIR E 5 14 12 CIR E-L 5 21 21 CIR E-C 2 7 6 CIR F 1 3 3 CIR F-C 1 3 3 FDR E 1 4 4 FDR E-C 1 4 4 FDR F-C 4 15 12 Asphalt Base Mixture 3 9 9 TOTAL 26 89 81 E = emulsiï¬ed asphalt, F = foamed asphalt, C = hydraulic cement, and L = lime. Table 13. Test specimen summary.
RLPD Test Results 45 5.3 RLPD Data Envelopes After removing all outliers, a qualitative measure was taken to compare the permanent deformation performance of each mix type, primary stabilizing agent, and chemical addi- tive. RLPD envelopes were defined by the greatest and least deformation curves for each mix combination. Figure 34 shows the RLPD envelopes for CIR, CCPR, and FDR mixtures. Each of these combinations contains some mixes that are stabilized with foamed or emulsified asphalt and may contain a chemical additive. CCPR and CIR have a higher upper limit than FDR, meaning that they are subject to higher deformation. The CIR and CCPR have good agreement, with the lower limit of the CIR being below that of the CCPR. This would indicate that some of the CIR mixtures tested have better deformation properties. Only three CCPR mixtures are represented in this data set, however, whereas 14 CIR mixtures are represented. It is possible that with more CCPR mixtures even more agreement would occur between CIR and CCPR mixtures. The FDR envelope shows the least deformation, meaning that it has the least deformation. When comparing these results to the dynamic modulus data trends represented in Figure 15 and Figure 16, similar relationships can be found at the lower frequencies. The FDR mixtures had the highest dynamic modulus values, a result that corresponds well with the lowest per- manent deformation characteristics from the RLPD testing. Similarly, the CCPR mixtures had slightly lower dynamic modulus values than did the CIR mixtures, which corresponds to the slightly higher rate of permanent deformation in the RLPD data. When compared to that of the CCPR mixtures, the dynamic modulus envelope of the CIR mixtures had a greater upper bound. This result similarly corresponds to a lower bound deformation found in the RLPD envelope. Figure 35 shows the same data seen in Figure 34 but adds the RLPD results of three asphalt base mixtures for comparison. Figure 35 shows that the base asphalt mixture envelope overlaps Figure 34. RLPD envelopes for mixtures produced by FDR, CIR, and CCPR.
46 Material Properties of Cold In-Place Recycled and Full-Depth Reclamation Asphalt Concrete Figure 35. RLPD envelopes for mixtures produced by FDR, CIR, and CCPR, as well as an HMA base mixture. with much of the area covered by the CCPR and CIR envelopes. This overlap indicates that the permanent deformation characteristics of CCPR and CIR mixtures may be similar to those of traditional base asphalt mixtures, whereas FDR mixturesâwhose RLPD envelope was lower than those of the other mixturesâmay have better performance. This finding complements ongoing studies that show good rutting performance of recycled mixtures, similar back- calculated stiffness values from falling weight deflectometer (FWD) data when compared to base asphalt mixtures, and similar dynamic modulus test results (Jenkins et al. 2007, Diefenderfer et al. 2016). The RLPD envelopes comparing emulsified asphalt to foamed asphalt are shown in Figure 36. As with the dynamic modulus data, no clear distinction is seen between the recycling agent types. The emulsified asphalt has a slightly higher upper limit than does the foamed asphalt, but the lower limit of the foamed asphalt is much lower than that of the emulsified asphalt. Fig- ure 36 represents 17 emulsified mixes and six foamed mixes. When comparing the dynamic modulus envelopes to those presented in Figure 18 and Figure 19, a similar trend in stiffness can be found. Generally, the mixtures containing foamed asphalt exhibited a higher stiff- ness value at the lower frequencies. The lower limit of the dynamic modulus values for the mixtures containing emulsified asphalt was slightly lower than that for the foamed asphalt, which corresponds well with the slightly higher deformation characteristics of emulsified asphalt seen in Figure 36. The data envelopes showing the results of chemical additives are shown in Figure 37. The pres- ence of cement yielded both the highest and the lowest permanent deformation envelope limits. This same trend was found in the dynamic modulus data shown in Figure 20 and Figure 21. The lower envelope limit of the recycled mixtures with no chemical additive was the highest,
RLPD Test Results 47 Figure 36. RLPD envelopes for mixtures produced using emulsified asphalt and foamed asphalt as stabilizing/recycling agents. Figure 37. RLPD envelopes for mixtures produced with lime, with cement, or with no additive.
48 Material Properties of Cold In-Place Recycled and Full-Depth Reclamation Asphalt Concrete indicating that these mixtures consistently experienced more permanent deformation. The mix- tures containing lime performed better than the mixtures with no chemical additive, with some overlap of the data envelopes near the lower limit of the mixtures with no chemical additive. The addition of cement was found to greatly decrease the permanent deformation characteristics in the RLPD test. Similarly, the dynamic modulus data showed an increased modulus for mixtures containing cement. 5.4 Relationship between Rutting Susceptibility and Density To investigate the relationship between rutting performance and density, the measured bulk density versus slope and intercept values were plotted for each specimen tested. For all in-place recycling processes after outliers were removed, Figure 38 shows the slope versus density and Figure 39 shows the intercept versus density. No clear trend could be established between the slope or the intercept in relation to density. This relationship was further investigated by primary stabilizing agent and by the presence of a chemical additive. In both cases, no clear relationship was found between density and RLPD output. Figure 38. Relationship between bulk density and RLPD slope for CIR, CCPR, and FDR mixtures.
RLPD Test Results 49 Figure 39. Relationship between bulk density and RLPD intercept for CIR, CCPR, and FDR mixtures.
