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.
34 The objective of this research was to use accelerated pave- ment testing techniques to conduct the rutting, fatigue, and moisture susceptibility validation experiments identified in NCHRP Project 4-19 by Kandhal and Parker (1) as shown in Table 1. The validation effort involved subjecting HMA mix- tures prepared with various aggregates to full-scale accelerated pavement testing and measuring their performance according to one of three HMA failure modes: (1) rutting; (2) moisture susceptibility; and (3) fatigue. The results are discussed below and are summarized in Table 26. Coarse Aggregate Uncompacted Voids Content Rutting The coarse aggregate UVA was found to be the best single predictor of rutting performance of the coarse-graded mix- tures as indicated by the descriptive ranking. The test appears to capture information related to particle shape and texture and rutting decreases as the coarse aggregate UVA increases. As discussed in Chapter 3, the coarse aggregate UVA appears to be less important at lower traffic levels. The number of APT wheel passes to reach a rut depth of 3.5 mm covers a range of approximately 100. The coarse aggregate UVA is more sensi- tive at higher traffic levels where the number of wheel passes to reach a 7-mm rut depth covers a range of approximately 14,000. In both cases, the relationship between traffic and coarse aggregate UVA seems less sensitive for UVA values in the range of 40 to 45 percent. The relationship becomes stronger in the coarse aggregate UVA range of 45 to 50 per- cent. Previous testing in the APT has indicated that one APT pass is equivalent to approximately 2,500 ESAL. Applying this equivalency factor to the coarse aggregate UVA/wheel pass data, a performance limit occurs at 100,000 ESAL. For expected traffic below 100,000 ESAL, a minimum coarse aggregate UVA of 40 percent would be required. A coarse aggregate UVA of at least 45 percent would be required for traffic above 100,000 ESAL. The preceding discussion of coarse aggregate UVA applies to the coarse aggregate UVB as well. The UVB test results had a high descriptive ranking as did the UVA test results. Thus, either test can be used to specify coarse aggregates used in HMA mixtures. Fatigue The research shows the trend that, as the coarse aggregate UVA increases, HMA mixture resistance to fatigue cracking increases. Although this trend can easily be seen in the data and appears logical, it is based on only three data points and other factors contribute to HMA mixture fatigue perfor- mance, such as initial mixture density, binder type, and pave- ment cross section. Nevertheless, there is no logical reason to discard the coarse aggregate UVA test as being unrelated to mixture fatigue performance. Flat or Elongated Particles Rutting The percentage of flat or elongated particles, 2:1 ratio, does exhibit a predictive relationship with HMA mixture rutting performance. Rutting increases with increasing FOE21. The FOE21 has a descriptive ranking only slightly lower than coarse aggregate UVA. In fact, FOE21 was found to be posi- tively correlated with UVA (i.e., as FOE21 increases, so does UVA). The two tests seem to predict HMA mixture rutting equally well. However, when the two descriptors are com- bined, they do not improve rutting prediction. If only one of the two tests is to be used, the coarse aggregate UVA seems preferable because it is typically less time consuming than the FOE21 test. However, given that the FOE21 and coarse aggre- gate UVA results are positively correlated, it would be good C H A P T E R 4 Conclusions and Suggested Research
35 practice to include both tests when specifying coarse aggre- gates for use in HMA. If the coarse aggregate UVA test is used alone, aggregates with very high FOE21 values could be used because they would also have high UVA values. Aggregates with extremely high FOE21 values could cause problems in HMA mixtures. Most coarse aggregates used in this research have FOE21 values between 35 and 50 percent. The FOE21 value of 50 percent appears to be a reasonable upper limit for specifica- tion purposes. Fatigue The research shows that as FOE21 increases, HMA mix- ture resistance to fatigue cracking increases. It should be recognized that this trend is based on only three data points and that many other factors contribute to the fatigue per- formance of an HMA mixture. Nevertheless, there is no log- ical reason to discard the FOE21 test as an aggregate test related to fatigue performance. Fine Aggregate Uncompacted Void Content Rutting Rutting performance of fine-graded HMA mixtures is related to fine aggregate UVA. The HMA mixture rutting resistance increases with increasing fine aggregate UVA. Both fine aggregate UVB and VTM5 test results have slightly higher descriptive rankings than the UVA results. As a result, the fol- lowing discussion of fine aggregate UVA applies to both fine aggregate UVB and VTM5 methods. It seems appropriate to recommend that any of the three tests can be used for speci- fication purposes for fine aggregates used in HMA mixtures; the choice can be left to the specifying agency. Similar to the coarse aggregate, the relationship between traf- fic and fine aggregate UVA appears to be less sensitive at lower traffic levels than at higher levels. In the case of fine aggregate UVA, the value of approximately 45 percent seems to be a logi- cal break point. HMA performance is markedly improved with fine aggregate UVA values above 45 percent compared to HMA with fine aggregate UVA values below 45 percent.Using the APT equivalency factor of 2,500 ESAL per pass, it might be appro- priate to allow the use of fine aggregate UVA as low as 40 per- cent for traffic levels below 500,000 ESAL and 45 percent or higher for traffic levels of 500,000 ESAL or higher. The results of APT moisture susceptibility tests indicate that, in the presence of moisture, fine aggregate UVA is a bet- ter predictor of rutting performance than any other test; rutting decreases with increasing fine aggregate UVA. The sensitivity discussion previously applied to dry APT testing appears to be true of tests in the presence of moisture as well. Fatigue Data indicate a relationship between HMA mixture fatigue performance and fine aggregate UVA might exist. As the fine aggregate UVA increases, HMA mixture fatigue cracking Proposed Specification Test Recommended for Specification Use Traffic (ESAL) Limit Less Than 100,000 40%, MinimumUncompacted Void Content of Coarse Aggregate, Method A (AASHTO TP56)1 Yes 100,000 and above 45%, Minimum Flat or Elongated Particles in Coarse Aggregate (ASTM D4791) Yes All 50%, Maximum Less Than 500,000 40%, MinimumUncompacted Void Content of Fine Aggregate, Method A (ASTM C1252)2 Yes 500,000 and above 45%, Minimum Methylene Blue Test for Fine Aggregate (AASHTO TP57) No â â Methylene Blue Test for p0.075 Material (AASHTO TP57) No â â Particle Size Analysis of p0.075 Materials for Determining D60, D30, and D10 Sizes No â â Micro-Deval Test (AASHTO TP58) Yes All 15%, Maximum Magnesium Sulfate Soundness Test (AASHTO T104) Yes All 20%, Maximum 1 Same recommendation and proposed specification applies to Method B as well. 2 Same recommendation and proposed specification applies to Method B and Virginia Test Method for Determining Percent Voids in Fine Aggregate (VTM5) as well. Table 26. Test recommendations and proposed specification limits.
decreases. The trend, if it does exist would seem logical. How- ever, with the limited data obtained in the research (three data points) and given that two of the mixtures never exhibited fatigue cracking, the fine aggregate UVA cannot be recom- mended as an aggregate test related to fatigue performance. A relationship may exist, but additional testing is needed before a conclusive recommendation can be made. Methylene Blue Test Based on the AASHTO T 283 tests of cores taken from the APT test lanes, the MBV appears to be somewhat related to the TSR values. The higher the MBV, the lower the TSR value or the more susceptible the HMA mixtures are to moisture. How- ever, the relationship does not seem strong and appears to be affected by the amount of p0.075 material in the mixture. For example, mixture FAM1 had a high MBV, but low p0.075 con- tent. Despite the high MBV, the mixture had a high retained tensile strength after being subjected to one freeze-thaw cycle. Furthermore, research covered in the literature review indi- cates a high MBV does not always indicate the presence of harmful material in the aggregate blend. Research results suggest the MBV test is not a good predic- tor of HMA performance and should therefore not be used to specify HMA mixture fine aggregates. Particle Size Analysis The effect of the p0.075 fraction on fine-graded mixtures was investigated using the APT. Overall, the D60 and D30 val- ues show a fair correlation with rutting performance. HMA mixture total rut depth increases as the amount of D60 or D30 material increases. When used in conjunction with the fine aggregate UVA, neither the D60 nor D30 parameters improve the fine aggregate UVA prediction of HMA mixture rutting. The D10 parameter does not appear to correlate at all with HMA mixture rutting performance. Considering the effort that goes into completing the particle size analysis of the p0.075 fraction and its fair ability to predict mixture per- formance, particle size analysis of the p0.075 fraction is not recommended for routine aggregate specification testing. Micro-Deval and Magnesium Sulfate Soundness Kandhal and Parker (1) recommended MDEV and MGSO4 tests to address aggregate durability and toughness when used in HMA mixtures. The MDEV for both coarse and fine aggre- gates and the MGSO4 for fine aggregates were determined and correlations were made with rutting performance. The MDEV descriptive rankings for coarse and fine aggregates were 2.7 and 5, respectively. The descriptive ranking for the fine aggregate MGSO4 was 4.2. Coarse aggregate MDEV has little correlation to mixture rutting performance. The fine aggre- gate MDEV and MGSO4 have fair correlations with mixture rutting performance. If only data at 20,000 APT wheel passes is considered, the correlation between rutting performance and the fine aggregate MDEV and MGSO4 are excellent. The amount of rutting at 20,000 APT passes increases as the fine aggregate MDEV and/or MGSO4 values increase. However, the opposite trend exists at 1,000 passes. Nonetheless, there does appear to be a good enough correlation between fine aggregate MDEV and MGSO4 to recommend their use to specify aggregates used in HMA mixtures. A maximum value of 15 and 20 percent for MDEV and MGSO4, respectively, would limit rutting at 20,000 wheel passes to 12.5 mm. Summary The research has shown that for coarse aggregates, both UVA (UVB) and FOE21 are good predictors of HMA mixture per- formance. It is recommended the tests be used in all climates and for all materials. For traffic above 100,000 ESAL, a coarse aggregate UVA of 45 percent or greater is recommended. A minimum UVA of 40 percent is recommended for traffic below this level. The FOE21 value should be a maximum of 50 per- cent for all traffic levels. For fine aggregates, UVA, UVB, or VTM5 tests can be used to control HMA rutting performance. The MDEV and MGSO4 tests on fine aggregates can also be used to control HMA mixture rutting performance. The UVA (UVB, VTM5) should be used in all climates and for all materials. For expected traffic of less than 500,000 ESAL, the minimum recommended UVA (UVB, VTM5) value is 40 percent; a minimum UVA (UVB,VTM5) of 45 percent is recommended for expected traf- fic greater than 500,000 ESAL. The MDEV and MGSO4 tests should be used in all climates and for all materials. These tests may be important in high moisture climates. A maximum value of 15 and 20 percent for MDEV and MGSO4, respectively, are recommended for all traffic levels. Recommended Research Problems were encountered in the fatigue testing portion of the experiment because of the lack of temperature control. Also, only limited fatigue tests could be incorporated in the study because of budget and time constraints. As a result, per- formance relations in the study are not as strong as they need to be. Although it is true that fatigue performance depends on coarse and fine aggregates, other factors contribute to the per- formance, including gradation, binder type and volume, tem- perature, initial mixture density, pavement thickness, applied loads, tire pressures, and traffic speed and volume. Gradation, 36
37 binder volume, and binder grade and type are likely the most significant factors related to fatigue performance. The other items are of only slightly less importance. A comprehensive laboratory study was undertaken in NCHRP Project 4-19 to screen aggregate properties and tests related to HMA mixture rutting and moisture effects. The cur- rent study built on the experience of that research to produce positive results. It is recommended that a laboratory fatigue study be conducted with a goal of screening aggregate charac- teristics and tests affecting HMA fatigue performance. Nor- mally, the difference between laboratory and in-service fatigue performance is a matter of using a scaling factor. For this rea- son, the research team believes that adequate information can be collected without the need for further full-scale testing.
