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.
44 This chapter presents case examples of agencies that have dealt with compaction and thin lifts in a variety of ways. The early experiences of the Florida DOT with excessive permeability of coarse graded Superpave mixes and how they modified their specifications are summarized first. Next, South Carolinaâs use of thick lifts for rapid construction and full-depth patching is highlighted. This is followed by Marylandâs specifications for dealing with the other extremeâ thin lifts. FHWAâs national research on ways to increase asphalt pavement density is reviewed. Last, the thoughts of an experienced mix designer are offered to reflect some industry perspec- tives and approaches. 4.1 FloridaâThe Minerâs Canary? The Florida DOT may not have been the first to experience problems related to compacting early Superpave mixtures at existing lift thicknesses, but it was certainly the first to share its experience and raise the issue at a national level. In 1996, the Florida DOT began to see water âweepingâ out of coarse graded Superpave mixtures on one of its first major Superpave projects. When cores were cut using a dry method, the core holes immediately filled with water that had been trapped in the upper layers of the pavement. An asphalt rubber crack relief layer appeared to be preventing the water from moving down through the pavement and instead directed it to the low side shoulder, where a fine graded Marshall mix prevented further lateral movement. This prompted the Florida DOT to look at its other Superpave projects. It found that six out of eight were much more permeable than the previously used Marshall mixes. One project was relatively impermeable; it had been constructed with a 12.5-mm mix placed in two 50-mm lifts, or 4ÃNMAS. The projects where permeability was a problem had t/NMAS ratios of less than 3. This suggested that a t/NMAS ratio of 4 for coarse graded mixes could improve the compactibility and reduce the permeability of Superpave mixes. Based on this examination, the Florida DOT increased its lift thickness to 4ÃNMAS for coarse graded mixes. Where fine graded mixes were used, the ratio was close to or equal to 3. When Superpave was first implemented in the state, the Florida DOT required coarse graded mixes for high-traffic applications, greater than 10 million ESALs. For lower traffic volumes, the contractors had the option of using either coarse or fine mixes. In 2005, the specifications were changed to make it mandatory to use fine graded mixes for less than 10 million ESALs, while for traffic levels greater than 10 million ESALs it was still mandatory to use coarse graded mixes. In 2007 the specifications were changed again to make the type of gradation optional for traffic levels greater than 10 million. Traffic levels less than 10 million ESALs were still mandated to be fine graded. At that time the vast majority of contractors stopped using coarse mixes as there were too many density and constructability issues when using them. In 2014 the specifications C H A P T E R 4 Case Examples
Case Examples 45 were officially changed to require fine graded mixes for all traffic levels. This reflects the evolu- tion of the Florida DOTâs specifications as it gained experience with the new mix design system. These were not the only changes and improvements the Florida DOT was implementing; Figure 38 shows some of the other changes Florida DOT made to improve the statewide pave- ment performance. Florida has a statutory requirement that no more than 20% of the state highway system can be deficient. The Florida DOT also changed how it measured density for acceptance in 1997. When first using Superpave in 1996, density acceptance was based on nuclear density readings; the gauges were correlated to cores during construction of a control strip. When the permeability issues sur- faced, forensic analysis showed that the in-place density was substantially lower than the gauges indicated (about 88% Gmm versus 92% to 93%). Core densities were more accurate, so they were used for acceptance with a target density of 91%. Contractors could use gauges for quality control. Some time later, the Florida DOT tightened its density requirements to 94.0% of Gmm when it still saw issues with permeability of coarse graded mixes at 91% density. This was based on the finding that even coarse Superpave mixes were âvirtually impermeableâ when the air void con- tent was less than 6%, but as the air voids increased above 7%, permeability increased rapidly. Experience has shown that this density could be attained with the increased lift thickness and attention to detail. Figure 38. Florida statewide pavement performance and timeline of changes (courtesy of Jim Musselman). 0% 2% 4% 6% 8% 10% 12% 14% 16% 18% 20% 22% 24% 26% 28% % D ef ic ie nt PCS Year Crack Ride Rut Adopted Superpave Improved Training Courses Adopted Contractor QC Adopted PWL Specs Adopted Warranty Specs Polymer modified binders Fine graded mixes Reduced resurfacing program 2002 1994 1995 1996 1997 1998 1999 2000 2001 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
46 Impact of Asphalt Thickness on Pavement Quality In addition to being willing to make specification changes to address the issues it was seeing, the Florida DOT was also willing to share its experiences with other agencies. Part of that willing- ness stemmed from Floridaâs role as one of the AASHTO Superpave Lead States. The Lead State program was initiated to identify state champions for some of the new technologies developed during SHRP. The Superpave Lead State team and activities are described in the NCHRP report on the history of Superpave (McDaniel et al., 2011). It was also partly due to the desire to get input from others on how to address the problems; FHWA proved to be especially helpful in this case. Aside from communicating with other states, Florida has a long-standing practice of regular communication within the department to avoid the formation of siloes. For many years, there have been conference calls scheduled weekly with the district materials engineers and monthly calls with the district bituminous engineers. Beginning about 2008, regular calls have been held between materials staff and construction staff. Some materials staff members talk almost daily with project engineers and several times a week with district construction engineers. This fre- quent interaction benefits everyone and helps the pertinent parts of the agency become aware of issues so they can be dealt with promptly. 4.2 Through Thick and Thin There are situations when it would be beneficial to be able to place asphalt mixes in thick lifts, such as when constructing a full depth patch under heavy traffic. In fact, any time maintenance of traffic (MOT) issues call for rapid construction, there could be advantages to placing fewer, thicker lifts. However, there are also situations that call for thin lifts to be placed, such as when dealing with overhead clearance, curb height, or variable depth paving (e.g., wedge and level). This section offers two case examples of how some states deal with these difficult situations. The first is South Carolinaâs use of thick lifts for rapid construction. The second is Marylandâs specification for density control on thin lifts. 4.2.1 Thick Lifts in South Carolina The South Carolina DOT has dealt with several projects where heavy traffic demands have forced the department to consider options to increase the speed of construction. Placing and compacting multiple lifts of asphalt in a deep patch or an overlay can create serious construction delays. Therefore, the South Carolina DOT has explored compacting deeper lifts than usually recommended. The South Carolina DOT has constructed a few interstate projects and has sponsored a test section at the NCAT Test Track using lifts up to 7 in. (175 mm) thick. One project on I-85 has an average daily traffic level of about 90,000 vehicles per day with 30% trucks. Traffic control was restricted to nightly and extended weekend lane closures. Traffic would be placed on the new asphalt the next morning, so a strong, rut-resistant mix that could be placed, compacted, and cooled quickly was needed. The final design chosen to correct cracking and other observed distresses was to mill 10 in. in the right-hand lane and 5 in. in the middle and left-hand lanes. The right-hand lane was repaved with two lifts at 495 pounds per square yard (psy) (4.5 in., 114 mm). The other two lanes were paved with one 450 psy (4 in., 100 mm) lift. The mix for this application was designed to be durable and easier to compact. The design was based on a 12.5-mm Superpave mix, but the design air void content was lowered from 4% to 2.5% and design gyrations decreased from 100 to 75. To maintain the stiffness, a PG 76-22 binder was used in lieu of a PG 64-22 and up to 30% RAP was allowed. A chemical warm mix asphalt technology was added to reduce the mix temperature and thereby reduce the mat temperature when opened to traffic. This modified mix is called âIntermediate B Special.â The original 12.5-mm Intermediate B mix had typically been placed in 2-in. lifts.
Case Examples 47 A test strip was required to ascertain if adequate compaction could be achieved. Full depth cores were pulled from the test strip and correlated to nuclear and nonnuclear gauges. The gauges were used to monitor compaction on the mainline without coring because coring would have delayed the construction process. The changes made in the mix design created a tender mix. This type of mix needs to be placed in confined conditions, like a mill and fill, because of the tenderness. Nonetheless, the contrac- tor was able to achieve density. The mix was under traffic for about 1 year before the surface was placed. No rutting or other distresses were observed and now, after more than 2 years, the pavement still looks good. Other projects placed on I-20 and I-385 have also performed well with lifts placed up to 595 psy (5.4 in., 137 mm). Cores pulled from some of these projects were cut in half and analyzed to determine if uniform compaction was achieved throughout the depth of placement, and the results were favorable. Placing fewer lifts does reduce the opportunities to improve smoothness. On one project, the IRI was acceptable at 80 to 90 before the surface was placed and less than 65 after the surface was paved. If the ride quality is substandard on other projects, diamond grind- ing may be considered at some point in the future. The South Carolina DOTâs special provision for full depth patching has been revised based on the success of these thicker lift applications. For patches less than 6 in. deep, either a surface or an intermediate mix is placed in two lifts. For depths of 8 or 10 in., contractors have the option of placing the Intermediate B Special mix in two lifts or a conventional Intermediate C mix in three lifts. Patches are required to be placed and opened to traffic the same day. To investigate how deep lifts could be, the South Carolina DOT sponsored a test section at the NCAT Test Track that was placed on August 23, 2018. A 12.5-mm mix was planned but the produced mix was actually a bit finer than desiredâreportedly being closer to a 9.5-mm mix. The mix was placed up to 7 in. (175 mm) deep. The rolldown observed during compaction was the typical Â¼ in. per inch of placement depth, or about 2 in. (50 mm) in this case. Results are preliminary, and data are still being analyzed, but indications are that the placement and com- paction were successful. In the words of one South Carolina DOT engineer, âthe 7-in. lift worked great.â NCAT will be reporting on the results in the near future and during traffic loading over the next few years. 4.2.2 Thin Lifts in Maryland In 2010, the Maryland State Highway Administration (MSHA) became aware of a number of cases of variable depth paving (such as wedge and level) or other circumstances where the lift thick- ness was thinner than typically desired. Contractors were having difficulties meeting the normal density requirement of 92% to 97%. To avoid a density disincentive, contractors were sometimes over-rolling the mat and breaking aggregates. MSHA decided that some way to define what a thin lift was and whether adequate density was being achieved was needed. So, a specification was drafted and then refined through partnering between MSHA and the Maryland Asphalt Association. A table (Table 1) was added to the specifications to define what constitutes a thin lift. For thin lifts, density is not measured in terms of % Gmm. Instead, the optimum density is determined by rolling a 400- to 500-ft. control strip. Density is monitored by thin lift nuclear or nonnuclear gauges. Optimum density is achieved when an additional roller pass does not increase the aver- age density by more than 1.0%. In addition, the optimum density is required to be at least 90.0% Gmm. Five cores are taken in the control strip to correlate the gauges, but coring is not done during regular construction (except when larger quantities of mix are placed in one loca- tion) unless changes in conditions or the mix indicate the need for a new optimum. Gauge
48 Impact of Asphalt Thickness on Pavement Quality readings are used to monitor compaction during paving. There is no incentive pay for density in these thin lift applications. Both industry and the agency are reportedly pleased with the specification. Contractors are able to compact the mix without âpounding the mix to death.â Involving the industry in the refinement of the specification made them receptive to the changeâand it solved a problem they were facing. The concept is also used in some other applications, such as when poor base or subbase conditions prevent achieving the usual density. In those cases, the agency acknowledges the optimum density is the best that can be achieved even though it may not reach 92% of Gmm. It is also sometimes used in residential areas where there are utilities or unconfined edges, or when vibratory compaction cannot be used. 4.3 Increasing Field DensityâA National Study The importance of achieving adequate density during construction is indisputable, and numerous studies have demonstrated that the ratio of lift thickness to NMAS has an impact on the ability to compact an asphalt mixture. However, there are many factors that can affect the compactability of an asphalt mixture and, therefore, its ultimate performance. In light of the profound impact of field density on pavement performance, the Federal Highway Admin- istration initiated a demonstration project called âEnhanced Durability of Asphalt Pavements through Increased In-Place Densityâ in 2015 (Bukowski, 2015). Under this demonstration project, states voluntarily agree to construct a control section using typical mix design, construction, and compaction procedures. They also constructed one or more test sections using different techniques in an attempt to increase the in-place density. The states chose the techniques that they believed would work in their situation. FHWA requested that at least one of the test sections be built at little or no cost (increased passes of existing rollers, temperature, etc.). In the first phase of the project, 10 states from coast to coast were selected to participate. The methods used to increase the field density included the following (Aschenbrener et al., 2017, 2018): â¢ Adding rollers, modifying rolling patterns, or changing types of rollers; â¢ Making engineered adjustments to mixture design parameters (gyration levels, design air voids, gradation, etc.), which often included mixture performance tests (these engineered adjustments led to increased binder content); Gradation Classification Control Sieve Mix Design Target (% Passing) Mix Designation Fine Graded Coarse Graded 4.75 mm A thin lift is a specified pavement thickness <1 in. A thin lift is a specified pavement thickness <1 in. 9.5 mm When the 2.36 mm (#8) is â¥47%, a thin lift is a specified pavement thickness <1â in. When the 2.36 mm (#8) is <47%, a thin lift is a specified pavement thickness <1Â½ in. 12.5 mm When the 2.36 mm (#8) is â¥39%, a thin lift is a specified pavement thickness <1Â½ in. When the 2.36 mm (#8) is <39%, a thin lift is a specified pavement thickness <2 in. 19.0 mm When the 4.75 mm (#4) is â¥47%, a thin lift is a specified pavement thickness <2Â¼ in. When the 4.75 mm (#4) is <47%, a thin lift is a specified pavement thickness <3 in. 25.0 mm When the 4.75 mm (#4) is â¥40%, a thin lift is a specified pavement thickness <3 in. When the 4.75 mm (#4) is <40%, a thin lift is a specified pavement thickness <4 in. 37.5 mm When the 9.50 mm (â ) is â¥47%, a thin lift is a specified pavement thickness <4Â½ in. When the 9.50 mm (â ) is <47%, a thin lift is a specified pavement thickness <6 in. Table 1. Thin lift mix design identification table (MSHA, 904.04.07).
Case Examples 49 â¢ Using WMA as a compaction aid; and â¢ Using new technologies to assess the placement and compaction process, such as intelligent compaction and thermal imaging. Two states in the first phase had test sections allowing comparison of differing NMASs in the same lift thickness (thus yielding differing t/NMAS ratios); both were constructed in 2016. In one state, test sections were placed on a two-lane road with a design traffic level of 1 million ESALs in a mill and overlay operation. The control section and two test sections had t/NMAS ratios of 3.0 for the surface course using a Â½-in. (12.5-mm) mix. Three test sections had surface courses with t/NMAS of 4.0 by using a Â¾-in. (9.5-mm) mix at the same lift thickness. Grada- tion information is not available for this project, so it is not known if these were coarse or fine mixes. Other test sections included changes in the number and types of rollers, changes in the binder content, and use of WMA. The average in-place densities of the six sections (control plus five tests) on this roadway were greater than 93.6%, and five of the six were between 93.6% and 94.2% of Gmm. The overall average density for the 12.5-mm surface at a t/NMAS of 3.0 was 94.1%, and the average for the 9.5-mm mix (at 4:1) was 93.8%. Given the narrow range of rela- tively high densities, the effect of changing the NMAS was not significant, probably because the density was already high (Aschenbrener et al., 2017, 2018). In another state, the different mix sizes were used on a two-lane road with a design traffic level of 2 to 8 million ESALs. Both mixes were described as âslightly on the fine side of the primary control sieveâ (Aschenbrener et al., 2017). The 12.5-mm mix used in the control sec- tion and six of the test sections had a t/NMAS of 3.5 while the ratio for a 9.5-mm mix used in one test section was 4.7. (Other test sections included various combinations of compactive effort, binder content, use of WMA, and mix temperatures.) In this case, the difference in the average densities was significant. The 12.5-mm mix control section had an in-place density of 93.5%. The control and six of the test sections had an average in-place density of 94.1% of Gmm. The test section with the 9.5-mm mix had an average density of 95.2. The density on the control section was 1.7% lower than that in the 9.5-mm test section. Recall that previous research shows that a 1% reduction in air voids (a 1% increase in density) can increase the service life by 10% (Tran et al., 2016). This somewhat limited comparison suggests that it is possible to achieve reasonably high densities with attention to detail, and a wide variety of techniques can help in the effort. While a greater t/NMAS ratio can help to increase density in many cases, it may not help in every case depending on other properties of the mix, pavement structure, construction operations, and other variables. Keep in mind, however, that none of the t/NMAS ratios evaluated here were under 3.0, which is the commonly recommended minimum for fine graded mixes. A second phase of this demonstration project involved nine more states; a report of this phase is expected by mid-2019. A third phase will include 10 more states with construction in 2018 and 2019. At the current time it is unknown if any of the states will consider an evaluation of differing lift thicknesses on these new construction projects. FHWA may revisit these projects to observe the performance after several years, but this effort is not currently funded. 4.4 A Contractor Seeks Solutions Many, if not most, contractors have had to make modifications in their materials, mix designs, or practices to attempt to meet or exceed density requirements. One contractor was willing to share his experiences as to what works and what further changes would be helpful. He sees the issue of lift thickness as being vitally important to the industry since it has such a great effect on the ability to obtain adequate compaction.
50 Impact of Asphalt Thickness on Pavement Quality This contractor has made a number of changes to their practices, such as using finer mixes and greater compactive effort. The Bailey method has been used to guide changes in the aggregate gradations to obtain dense, stable mixes while allowing enough room for an adequate film thick- ness. The Bailey method seems to be especially helpful with coarse graded mixes. While these changes have been largely successful, this contractor sees more room for improvementâbut these changes would need to be allowed or initiated by the agencies. Specifically, this mix designer advocates for the use of thicker lifts. He agrees with the NCHRP Report 531 recommendations of a minimum of 3ÃNMAS for fine graded and 4ÃNMAS for coarse graded mixes. Based on field experience during his career, he also believes there are maxi- mum ratios as wellâat 6ÃNMAS for fine and 8ÃNMAS for coarse mixes. However, he sees that agencies and owners typically specify lift thicknesses at or near the minimum for economic reasons. Utilizing lift thicknesses in the middle of the ranges suggested above (i.e., 3ÃNMAS to 6ÃNMAS for fine graded, and 4ÃNMAS to 8ÃNMAS for coarse graded) would improve field compactability. The perception seems to be that agencies simply do not have the funds to increase lift thick- ness. While budgets are limited, there are ways to accommodate increased lift thicknesses with- out substantial budgetary impacts. âWe donât have to place the same mix thicker, we can utilize the next smaller NMAS at the originally planned lift thickness,â he says. As observed in the lit- erature review, smaller NMAS mixes are not necessarily weaker or less stable than larger NMAS mixes, especially with the other changes Superpave mix design brought, such as higher aggregate angularities, less natural sand, and polymer modified binders. Costs will still increase as VMA and asphalt content are tied to NMAS. However, there will be a notable increase in density for an equal compactive effort being applied, a decrease in permeability, and an improvement in mix performance.