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.

233 mixes, the authors infer that the mixture with PG 76-34 was still resistant to moisture damage although the actual tensile strength was much lower than expected. 1.59.8 Structural Design No information is provided on structural design. 1.59.9 Limitations No information is provided on limitations of use. 1.60 Watson, D. E., E. Masad, K. A. Moore, K. Williams, L. A. Cooley, Jr. “Verification of VCA Testing To Determine Stone-On-Stone Contact of HMA Mixtures.” Transportation Research Record No: 1891. Transportation Research Board. National Research Council. Washington, D.C. 2004. 1.60.1 General In this paper, Watson et al describe a laboratory study evaluation of stone-on-stone contact in Open Graded Friction Course (OGFC). Working on the premise that stone-on- stone contact is important for the proper functioning of OGFC, the authors conducted this study to determine whether stone-on-stone contact exists in mixes with different aggregates (same gradation, but using different compaction methods and levels), verify the existence of stone-on-stone contact in these mixes with digital imaging techniques and also to determine the effect of aggregate breakdown (during compaction) on stone- on-stone contact. Watson et al considered the voids in coarse aggregate in mix (VCAMIX) being less than the dry-rodded voids in coarse aggregate (VCADRC) as the condition to determine existence of stone-on-stone contact. Using the percent retained on the 4.75 mm sieve as the coarse aggregate fraction (breakpoint sieve), the authors have shown that for this specific 12.5 mm Nominal Maximum Aggregate Size (NMAS) mix, only one of the fifteen different mixes did meet the VCAMIX<VCADRC criterion. On the other hand, they found all but one mix met the criterion, when the 2.36 mm sieve was used as the “breakpoint” sieve. Watson et al. made a recommendation that the finest sieve to retain more than 10 percent of aggregates be considered as the “breakpoint” sieve for determination of stone-on-stone contact. With X-Ray computed tomography images, Watson et al. used three different image analysis techniques to verify stone-on-stone contact in OGFC mixes: number of contacts, level of contacts, and air void size distribution. The authors found reasonable to good correlation between the results of the three methods. The general conclusions were that there was more stone-on-stone contact in Marshall compacted specimens than in Superpave gyratory compacted specimens and an increased level of compaction did increase the number of “contacts” for mixes with all three aggregates except one. Watson et al indicate that the amount of aggregate breakdown was higher for Marshall compaction than for Superpave gyratory compaction. They suspect that the higher breakdown caused greater number of intermediate particles and lowered the VCAMIX, or

234 that the lower compaction effort was not enough for locking the coarse aggregate particles together. In general, Watson et al concluded that the finest sieve to retain more than 10 percent of aggregates can be considered as the “breakpoint” sieve for determination of stone-on- stone contact. They also indicate that digital image analysis can provide an independent method of verifying stone-on-stone contact in mixes. 1.60.2 Benefits of Permeable Asphalt Mixtures No information is provided on benefits of permeable mixes. 1.60.3 Materials and Design Watson et al used three different aggregates, two different compaction techniques and several compaction levels. The materials and matrix of mixes are shown in Table 113. Table 113: Materials and Mixes Material/Property Type/Test Results Asphalt Binder Polymer modified PG 76-22 Asphalt Binder content, % 6.0 Aggregate Granite, Gravel, and Traprock Bulk Specific Gravity (Range) 2.599 to 2.927 LA Loss, % (Range) 17 – 36 Gradation Sieve Size, mm Percent passing 19 100 12.5 90 9.5 45 4.75 16 2.36 7 1.18 6 0.6 5 0.3 4 0.15 3 0.075 2.5 Additive Fiber Additive content 0.3 % of total mix weight Compaction method Superpave gyratory and Marshall Compaction level for Marshall 25, 50 Compaction level for Superpave gyratory 30, 45, 60 1.60.4 Construction Practices No information is provided on construction practices. 1.60.5 Maintenance Practices No information is provided on maintenance practices

235 1.60.6 Rehabilitation Practices No information is provided on rehabilitation practices 1.60.7 Performance Watson et al evaluated the performance of the different mixes, in terms of stone-on-stone contact, as determined from VCA calculations and digital image analysis. The paper presents results of volumetric calculations to determine the existence of stone-on-stone contact (VCAMIX<VCADRC), results of digital image analysis in terms of number of contacts, level of contacts and air void distribution, and discussion on the results from the analysis of volumetric and image data. Watson et al indicate that for Marshall compaction, the VCAMIX decreased with an increase in compactive effort. However, using the 4.75 mm sieve as the breakpoint sieve, the results showed that that only the gravel mix at 50 blows met the requirements of having VCAMIX < VCADRC. The authors speculate that this may be because the rounded gravel particles would more easily move past each other during the compaction process. Similar results were found for mixes compacted with the Superpave gyratory compactor, but in this case, none of the three mixes for the 4.75 mm breakpoint sieve met the requirements of VCAMIX < VCADRC. The authors mention that further reduction of percent passing the 4.75 mm sieve (to lower the VCAMIX) was not possible since the mix already had only 17 percent passing the 4.75 mm sieve. Also, the authors conclude that it is not the aggregate breakdown that is preventing the mixes from meeting the VCAMIX<VCADRC criteria, since the traprock aggregates, which showed the lowest breakdown, had the highest difference between the VCAMIX and VCADRC. Working on the assumption that most of the mixes did not meet the VCAMIX criteria, not because the criterion was not applicable, but because the breakpoint of 4.75 mm should be different, the authors show the VCAMIX and VCADRC data for the 2.36 mm sieve (as breakpoint sieve). With the 2.36 mm sieve as the breakpoint sieve, all of the mixes except for the traprock 25-blow Marshall mix, were found to meet the VCAMIX < VCADRC . Hence, the authors conclude that the 2.36 mm sieve is a better choice for breakpoint sieve for this OGFC mixture. In digital image analysis, Watson et al provided the average number of contacts for all the images captured within a specimen per unit volume (between 100 to 150 images/specimen). The packing parameter is calculated from the “Level of Contact Regions” method. A low packing parameter indicates that the particles have a strong level of contact (large area of contacts) and they are more difficult to be separated by the erosion operation compared with the ones that have higher packing parameter. The authors provide the ranking of the specimens according to the size of air voids at which 25 percent, 50 percent, and 75 percent of total air voids are smaller. Based on the hypothesis, that smaller air void size indicates denser packing, the authors mention that the gravel mixes at 50 Marshall blows showed the highest number of contacts or packing. The authors mention that a comparison between the number of contacts method showed an increase of about 30 percent in the number of contacts for gravel and traprock

236 specimens compacted to 60 gyrations compared with the ones compacted to 30 gyrations. There was almost no difference between the SGC granite specimens. The Marshall specimen at 50 blows had 20 percent more contacts than the Marshall specimen at 25 blows. The Marshall specimen at 50 blows had about 210 percent more contacts than the SGC specimen at 60 gyrations. Watson et al mention that the second method of using the level of contact region showed that the packing level increased by about 60 percent for the gravel SGC specimens compacted to 60 gyrations compared with the one compacted to 30 gyrations. Also, the packing level increased by about 60 percent for the Marshall specimens at 50 blows compared with the SGC specimens at 60 gyrations. This method, however, predicted that the level of packing decreased with increasing the number of gyrations for the granite and traprock specimens. For the air void distribution method, Watson et al indicates that there was a slight difference in packing for granite and traprock specimens compacted at 60 and 30 gyrations. However, the difference in packing was higher between the SGC gravel specimens compacted at 60 and 30 gyrations. Also, the packing of the Marshal specimens at 50 blows was about 30 percent to 50 percent more than the packing in the SGC specimens at 60 gyrations. The authors mention that this may be caused by the additional aggregate breakdown of the Marshall compaction versus the SGC. There was almost no difference in packing between the Marshall specimens at 25 and 50 blows. The authors conclude that in general, the Marshall specimens had more contacts than the SGC specimens and that the gravel specimens were more influenced by the increase in the number of gyrations from 30 to 60 than the traprock and granite specimens In their discussion, Watson et al point out that the fact that the granite mix compacted with the 25-blow Marshall hammer only marginally met requirements seems to indicate that the 25-blow Marshall method may not provide enough compaction energy to adequately lock the coarse aggregate particles together. They also indicate that the digital image analysis procedure showed stone-on-stone contact in samples of all mixes and, hence, the concept of VCA is applicable, but the concept of using the same breakpoint for all gradations is not. They indicate that the slope of the gradation curve should be a factor in making that decision. For breakpoint sieve, the authors provide a recommendation of using the finest sieve size for which there is at least 10 percent of the total aggregate retained. From the summary of the image analysis results, Watson et al. mention that a minimum number of contacts per specific area may be required to ensure stone-on-stone contact is obtained. For example, a minimum value such as 100,000 contacts per cubic meter may be recommended. They mention that this specification may be an advantage over the VCA method in that with digital imaging the number of contact points can be readily determined while the VCA method only gives a “yes” or “no” answer as to whether stone-on-stone contact exists. However, the authors do mention that more research is needed to compare aggregate packing from laboratory compaction to field compaction