Thin Asphalt Concrete Overlays (2014)

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.

10 chapter three DESIGN AND CONSTRUCTION OF THIN OVERLAYS Materials, mix design requirements, and construction proce- dures may have a significant effect on the longevity of thin overlays. Agency specification requirements and construc- tion practices provided through the questionnaire and sur- vey responses were used to identify some of the practices considered to be most useful for successful construction of thin overlays. MIX DESIGN The majority of state agencies (72%) use Superpave mix design procedures for designing thin overlays. Of the remain- ing agencies, the Marshall Mix design method described in Asphalt Institute MS-2 or an agency-specific design proce- dure is used (such as for UTBWC mix design). In keeping with Superpave criteria, most agencies design thin overlay mixtures in which the optimum asphalt binder content is based on 4.0% air voids. One of the early concerns with the Superpave mix design system is that implementation was encouraged based on mix- ture volumetric properties alone. Some agencies (Arkansas, Georgia, Idaho, Illinois, Massachusetts, Montana, New Jersey, North Carolina, Ohio, Oklahoma, Oregon, Pennsylvania, Texas, Vermont, and Washington) have added other perfor- mance criteria, such as a rutting test using either the APA or Hamburg Wheel-Tracking Device as a safeguard against possible rutting. A Texas study (Walubita and Scullion 2008) promotes the idea of a “balanced” mix design procedure in which maximum asphalt binder content is determined for which the mixture will be rut resistant, and a minimum asphalt content is needed to ensure resistance to cracking. The range of acceptable optimum asphalt content is within the range that the two parameters have in common. In the research, a Hamburg Wheel-Tracking Test is used to assess rutting potential and the Texas Overlay Tester (OT) is used to evalu- ate resistance to reflective cracking. The performance tests are conducted on samples at 7.0% ± 0.5% air voids to simulate typical in-place air void levels after construction. Since the implementation of Superpave in the 1990s, most agencies have found that rutting is no longer the major dis- tress to be dealt with but that cracking is of increas ing concern. Numerous research studies are being conducted in an effort to find a reliable test to predict resistance to cracking. Although several tests appear to be promising, currently there is no consensus on one test that appears to adequately and accurately relate laboratory performance to field performance. Transportation agencies, including those of Alabama, Georgia, Maryland, and others, have found that SMA can be expected to deliver superior performance to conventional dense-graded mixes even on high-traffic, heavy-load condi- tions. A 12.5-mm SMA mix has been the standard surface mix since the technology was brought to the United States from Europe in 1990. However, the German SMA surface mix is closer to a 9.5-mm NMAS. As agencies consider ways to improve performance with thinner layers of asphalt mix, it is only natural that interest in 9.5- and 4.75-mm SMA mixes has increased. The use of 9.5-mm SMA is routine in Alabama for high-traffic routes (Table 3). NCAT research (Cooley and Brown 2003) has recommended a gradation range for both 9.5- and 4.75-mm NMAS SMA mixes. MATERIAL REQUIREMENTS In general, agencies responded that they use the same material specification requirements for thin overlay asphalt mixtures as would be used for thicker layers. For example, if a Superpave mixture would normally be used, the Superpave specifications still apply, and if an SMA mixture is to be used, the same materials specifications that apply for a thicker SMA layer would be used. The only difference may be that a smaller NMAS mixture may be used. A Kansas study (Rahman et al. 2011) found that aggre- gate type also may affect performance. The study evaluated 4.75-mm mix placed between 0.6 and 0.75 in. (15–19 mm) thick; after 3 years, the mix composed of crushed gravel aggregates appeared to perform better than did the mixture with crushed limestone. Most agencies reported they would use the same asphalt binder grade for a project regardless of whether a thin or thick overlay was used. However, Kansas, New Jersey, Rhode Island, and Texas responded that they typically would use a different asphalt binder grade for thin overlays than if a thicker overlay were used on the same project. Most agencies responded that the decision to use modified asphalt usually was tied to traffic volume. For example, Kentucky uses mod- ified binder in surface courses for all interstate routes and roads with more than 30 million equivalent single axle loads

11 In the past, some agencies have been reluctant to use RAP, particularly in surface mixes. However, the economic advan- tages for doing so are clear. Missouri DOT reported that for the first 5 years after the agency began incorporating RAP in its asphalt mixes, the department had estimated savings of $34 million (Watson 2009). LABORATORY COMPACTION LEVEL (NDESIGN) Many agencies responded they use the Superpave-specified gyration level for laboratory compaction as specified in AASHTO M 323. Others have conducted research to evalu- ate the gyration level at which the aggregate structure begins to lock together. That gyration level then becomes the accepted gyration level for their asphalt mixes, especially for mix used in thin overlays. To achieve increased density beyond that point may fracture aggregate particles and create exposed aggregate faces that are uncoated and susceptible to mois- ture infiltration and stripping. Both Maryland and Georgia use 50 gyrations for compacting 4.75-mm NMAS mixtures used in thin overlays. Georgia, Maryland, and Virginia use 65 gyrations with a Superpave gyratory compactor as the NDESIGN level for other mixtures (including those placed on higher–traffic-volume projects), whereas Alabama has found the aggregate structure essentially locks at approximately 60 gyrations for the aggregate materials that agency uses. This practice generally allows a higher asphalt binder content. It is recommended that an agency determine the locking point of the aggregate structure in its mixtures and use that number of gyrations for its NDESIGN level, while keeping the binder type the same, especially for thin asphalt overlays. The locking point is defined as the first occurrence at which the specimen height remains the same for three successive gyrations (Watson et al. 2008b). The Georgia DOT study found that the locking point density correlated well with the ulti- mate density achieved under field conditions. The study also showed that mixtures designed at 60 gyrations had approx- imately twice the fatigue life as specimens for mixtures designed at 110 gyrations. TESTING CONSTRAINTS OWING TO SMALL NORMAL MAXIMUM AGGREGATE SIZE OR THIN LAYERS One of the concerns with standard laboratory testing for thin overlays is that the procedures and specification parameters were often developed for coarser mixes that are placed in thicker layers. For example, is it reasonable to assume that a 4.75-mm NMAS specimen prepared for AASHTO T 283 moisture susceptibility testing in a 6-in. (150 mm) diameter mold at a thickness of 3.75 in. (95 mm) will perform the same when placed on the roadway at 0.75 in. (19 mm) thick? A limited study for Georgia DOT compared the effect of Marshall samples with that of Superpave gyratory samples (ESALs) over a 20-year design life, and Montana uses modi- fied binders on all roads with more than 50 daily ESALs. Louisiana, New Jersey, Rhode Island, Utah, and West Vir- ginia reported using modified asphalt in all thin overlay sur- face mixes. One area that tends to be different for thin overlays is the use of recycled materials. Although the use of reclaimed asphalt pavement (RAP) and recycled asphalt shingle (RAS) is widely accepted for Superpave mixtures, many agencies do not allow RAP or RAS to be used in SMA and OGFC mixtures. Alabama is one of the few agencies that allow as much as 15% RAP in SMA mixes. Other agencies are begin- ning to consider RAP use in those specialty mixes. A 2008 study for Georgia DOT (Watson et al. 2008a) compared lab- oratory performance of four RAP types with four aggregate sources using four RAP proportions from 0 to 30%. The study found that virgin aggregate had a much greater influence on Los Angeles abrasion loss and percent flat and elongated particles than did the RAP aggregate. Higher RAP proportions increased mixture tensile strength values but reduced the fatigue life based on the bending beam procedure of AASHTO T 321. The study found that as much as 20% RAP could be used without significantly affecting performance. One con- cern with high RAP proportions is that the potential for low temperature cracking may be increased, but in the Georgia DOT study, as much as 30% RAP had little effect on the low temperature properties of the binder. Texas agency respondents indicated they are considering the effect of allowing RAP and RAS in SMA and fine-graded OGFC mixtures that may be used for thin overlays to 1 in. (25 mm) thick. Each mix was subjected to Hamburg Wheel- Tracking Test and OT laboratory performance analysis, and the study showed that these mixes perform exceptionally well. Results for rutting resistance increased as expected because of stiffness associated with the addition of RAP, yet the resistance to cracking as measured by the OT was still acceptable (Swaner 2012). Sieve (mm) 9.5-mm NMAS 4.75-mm NMAS 12.5 100 9.5 90–100 100 4.75 26–60 90–100 2.36 20–28 28–65 1.18 13–21 22–36 0.6 12–18 18–28 0.3 12–15 15–22 0.075 8–10 12–15 Source: Cooley and Brown (2003). TABLE 3 RECOMMENDED SMA GRADATION RANGES, PERCENT PASSING

12 150,000 tons of asphalt mix per year, a contractor could save approximately $75,000 annually if sheltered stockpiles reduced moisture content just 1% (Frank 2013). If RAP is used in production of fine mixes, it may need to be crushed or fractionated so that it meets the maximum size gradation requirements. Although crushing and sizing may not be required, it can be advantageous for a contractor to make the most efficient use of RAP in different NMAS mixtures. RAS material also will need to be shredded to a maximum size for the mix being produced. Production temperatures may need to be greater for thin overlays because they cool more quickly. When placed, an asphalt mixture begins cooling from both the bottom (exist- ing pavement temperature) and the top (ambient tempera- ture) so that thin layers have reduced time available for the on results for AASHTO T 283 (Watson et al. 2008b). Tensile strength of Marshall samples 4 in. (100 mm) in diameter × 2.5 in. (64 mm) thick were compared with gyratory samples 6 in. (150 mm) in diameter × 3.75 in. (95 mm) thick. A sta- tistical regression showed there was basically no significant relationship between the two results. Increases in tensile strength based on aggregate source with Marshall size speci- mens had a trend of decreasing strength for gyratory size samples (Figure 6). These results are of concern particularly because several of the mixture requirements for Superpave mixtures were copied from the mixture requirements developed for Mar- shall mixtures. Relating current test parameters, and possibly test procedures, to fit the sample sizes actually encountered in the field might be accomplished. For example, laboratory samples for tensile strength prepared according to AASHTO T 283 are 95 mm in thickness. A question arises as to correla- tion and applicability when those same parameters are used for roadway cores from a pavement layer 1.5 in. (38 mm) thick or less. PRODUCTION With fine mixes such as 4.75- and 9.5-mm NMAS mixtures typically used for thin overlays, plant production may be slower because material will need to be kept in the dryer for a longer period to remove moisture. The finer aggregate has more surface area and thus generally has higher moisture con- tent than does coarse aggregate. Private industry respondents have reported that using a storage shelter (Figure 7) for stock- piling fine aggregate, RAP, and RAS will soon pay for the investment with reduced drying costs. Plant diagnostic tools show that a 1% increase in moisture increases drying costs by approximately 10% to 12% while reducing production by approximately 11%. For a plant that produces approximately TS R6 TSR4 1401301201101009080706050 110 100 90 80 70 60 Scatterplot of TSR6 vs TSR4 FIGURE 6 Comparison of 4-in. Marshall versus 6-in. gyratory TSR results. (Source: Watson et al. 2008b.) FIGURE 7 Storage shed for aggregate stockpiles. (Source: Randy West, NCAT.)

13 CONSTRUCTION Surface Preparation For long-term service life with thin overlays, it is essential to resurface a candidate project that has a stable foundation with high severity distresses. Any areas of poor drainage need to be addressed before applying the overlay. The amount of sur- face preparation needed is dependent on the type and severity of the existing pavement distresses. Milling has several advantages when used on thin overlay projects. Milling helps to maintain existing grade so that bridge clearances and curb and gutter structures are not adversely affected. Milling equipment may use grade and slope controls to restore geometric shape and improve ride quality. A few states, such as Georgia, Ohio, Oklahoma, South Carolina, and Texas, have smoothness requirements for the milled surface. Those requirements are generally to eliminate isolated high or low spots, and measurements are taken with a straightedge. Georgia was the only state found to use inertial profiler mea- surements on a milled surface. Agencies need to have realistic expectations about how much improvement in smoothness can be obtained with just a thin overlay. A combination of milling and overlay may be used in such cases to meet smoothness contractor to obtain density. A 1-in. mat will cool twice as fast as a 1.5-in. mat in the typical compaction range of 300°F to 175°F (149° to 80°C) (Newcomb 2009). For this reason, contractors are finding that warm mix technologies are help- ful as a compaction aid, especially in cool weather, because they increase the working window of temperatures at which compaction may be accomplished. However, some agencies have concerns about very low production temperatures with WMA for fear the temperature may not be adequate to com- pletely remove aggregate moisture. Responding agencies realize that cooler temperatures will limit the capability of the contractor to properly compact thin overlays (Table 4). For that reason, most agencies limit placement based on layer thickness and ambient tempera- ture, and a few also have seasonal limitations. Mississippi, Alabama, and New Jersey allow lower temperatures if an approved warm mix technology is used. Mississippi reported a minimum ambient temperature of 55°F (12°C) is used for placement of thin layers, but for WMA, placement to 40°F (4°C) is allowed. New Jersey and Alabama normally require an ambient temperature of 45°F (7°C) for mix placement but will allow ambient temperatures to 35°F (2°C) when WMA is produced. State Minimum Ambient Temperature (°F) Comments State Minimum Ambient Temperature (°F) Comments AL >45 >35 for WMA NC >40 AK >40 ND >40 AR ≥55 NE ≥40 AZ >70 Surface temp >85 NH >50 CA > Freezing NJ >45 >35 for WMA CO >55 NM >60 DE >45 NV >45 FL <1 in. = 50; >1 in. = 40; >1 in. = 45 if PG > 76 or ARB-5; WMA = 5° OH <1 in. = 60; 1–1.4 in. = 50; 1.5–2.9 in. = 40 GA >45 >55 if < 1 in. & for OGFC OK >55 ID >60 (Surface Temp) OR >60 IL >45 PA ≥40 for 9.5 mm Seasonal calendar dates by region KS >55 RI >50 KY >45 SC >45 LA >50 TN >45 MA >45 TX >70 ME >50 Seasonal—May 15 to First Saturday after Sept. 15 UT >45 MD >45 VT ≥50 MO > Freezing WA > Freezing MS >55 >40 for WMA WV >50 MT > Freezing Source: Survey responses. Agencies listed by postal abbreviations. TABLE 4 MINIMUM TEMPERATURES FOR PLACING THIN OVERLAYS

14 agencies have some requirement in their specifications to address this issue, and many agencies even require a materials transfer vehicle (MTV) to be used for this specific purpose. An MTV typically has additional storage capacity so that the paving operation can continue even if no truck is present at the construction site. However, there have been projects even with an MTV where the paver was stopped for a significant portion of time because the contractor failed to provide an adequate supply of trucks (Figure 9). The MTV may also contribute to improved smoothness by keeping the delivery truck separated from the paver. For typical dense-graded mixes, responding agencies nor- mally specify a thickness at least three times the NMAS. Most agencies use these same criteria for placement of thin overlays. Florida, Georgia, Indiana, and Louisiana indicated they use 1.5 × NMAS for thin overlays, particularly those less than 1 in. (25 mm) thick. However, layer thickness is often dependent on mix type. For example, respondents of the Florida agency reported they typically use 3 × NMAS as a basis for layer thickness but use 1.5 × NMAS criteria for OGFC mixtures. Alabama, California, Idaho, Maryland, Minnesota, Montana, and Texas responded that 2 × NMAS is typically used for thin overlays. Surfacing with UTBWC may also be in the range of 1.5 to 2 × NMAS. South Carolina’s survey response cautioned that using placement rates too low may cause drag marks and other issues. One contractor also responded that placement rates were sometimes too thin for the NMAS being used and that an additional 25 lb/yd2 would help achieve density and smoothness, thereby providing longer life. Another contractor cautioned against using thin overlays on slow-moving traffic urban roads where such overlays do not perform as well in turning lanes and intersections. Compaction Density of the final mat may be difficult to determine, espe- cially for layers less than 1 in. (25 mm) thick owing to the process of coring and sawing the layer. The integrity of such a thin sample may be distorted so that results are not depend- able. For that reason, responding agencies report that they expectations. Milling also provides a rough surface texture that improves the bond between the existing pavement and the applied overlay. Another obvious advantage is that mill- ing provides a renewable resource that can be used to reduce production costs and provide a sustainable pavement. Cleaning the existing surface to remove dirt and silt, and applying a uniform tack coat is essential for helping to create bond between the newly applied surface mix and the exist- ing surface because the interface is so close to the shearing forces caused by traffic braking and turning on it (Hansen 2013). Nine agencies responding to the questionnaire identi- fied the amount of surface preparation as being one of the reasons for large variations in service life. If the tack coat is inadequate, slippage of the new overlay, particularly in braking areas, may be a significant problem (Figure 8). Based on survey responses, tack application rates vary from 0.02 to 0.2 gal/yd2 (0.1 to 0.96 L/m2). The rate varies depending on whether the overlay is dense-graded mix or UTBWC and whether the tack coat is emulsion, asphalt cement, or a special “trackless” type tack coat. The rate may also depend on the type of surface being placed and whether the existing surface has been milled. For example, Tennes- see applies 0.08 to 0.12 gal/yd2 on a milled surface and 0.05 to 0.1 gal/yd2 on a nonmilled surface; Kansas uses 0.03 gal/ yd2 for 4.75-mm NMAS mix and 0.13 gal/yd2 for UTBWC; and Louisiana uses 0.12 gal/yd2 for OGFC. If an emulsion is used for tack coat, the specified rate is usually based on the residual asphalt amount. Georgia found that slippage may be more of a problem when emulsions are used during hot sum- mer paving, and the agency has required asphalt cement for tack coat for the last 30 years. Placement With any asphalt mix construction quality is generally improved by maintaining a continuous operation. Most state FIGURE 9 Paving operation with use of MTV. (Source: Don Watson, NCAT.) FIGURE 8 Slippage of thin overlay. (Source: Don Watson, NCAT.)

15 Because smoothness is an acceptance parameter for many responding agencies, milling, leveling, or both may be needed to help meet the requirements when thin overlays are speci- fied. As a general rule, only 40% to 60% improvement in ride quality can be expected with a single layer of asphalt mix (Newcomb 2009). generally do not require a certain density level or target value for thin overlays. If the layer is less than 1 in. (25 mm), an agency may specify that the rolling effort must be to the sat- isfaction of the engineer. Other agencies may even specify the type rollers to be used and the number of passes to make. It is not uncommon for responding agencies to restrict the use of vibratory rollers in vibratory mode when compacting thin layers because of the potential of fracturing aggregate particles. Steel wheel rollers may be used, or vibratory roll- ers may be used in static mode. Pneumatic-tired rollers may also be required in an effort to improve density through the kneading action of the roller tires. For layers greater than 1 in. (25 mm), density is most often controlled by comparing results to the theoretical voidless density. Compaction may also be based on a percent of labo- ratory density or a percent of a field-established control strip. Some agencies reported specifying more than one method, depending on the mix type and/or layer thickness (Table 5). Several agencies use nuclear and/or nonnuclear density gauges to evaluate roadway density. Other agencies allow the contractor to use such gauges to monitor compaction during construction but use roadway cores for acceptance testing. Acceptance Criteria Responding agencies routinely use the same acceptance cri- teria for thin overlays as for thicker layers with the exception of density for layers less than 1 in. (25 mm) thick. Gradation, asphalt content, plant lab air voids, roadway density, and smoothness are commonly used criteria (see Table 6). Idaho, Illinois, and Mississippi also mentioned voids in mineral aggregate as an acceptance parameter. % of Control Strip % of Lab Density % of Theoretical Not Measured- Satisfy Engineer Not Measured- Rollers and/or Passes Specified GA DE AK AL AR ID MN CO KS AZ OH NH FL LA IN VT WA GA MS KS IL NC ME KY RI OK MA TN OR MD TX TN MO VT MT NC ND NJ NM NV OH PA VT WV Source: Survey responses. Agencies listed by postal abbreviations. TABLE 5 METHOD OF SPECIFYING DENSITY BY AGENCY Asphalt Content Gradation Plant Lab Air Voids Roadway Density Smoothness AK MO AK NC AL NC AK MT AK MO AR MS AR ND CO NE CA NC AR MS AZ NC AZ NH DE NH CO ND CA MT CA NE CA NM FL NJ DE NH CO NE CO NH CO NV ID NM FL NJ DE NH DE NM DE OH IL OR GA NM FL NJ FL NV FL OR IN PA IL NV GA NM GA OH GA PA KY RI KY OH KY NV IN PA IN RI MA TX MA PA IL OH KY RI LA SC ME VT MD WA IN OR LA SC MA TN MN WA MN WV LA PA MA TN ME VT MS WV MO MA WA ME WA MN WA MT MN WV MN MO MS Source: Survey responses. Agencies listed by postal abbreviations. TABLE 6 ACCEPTANCE CRITERIA FOR DENSE-GRADED THIN ASPHALT OVERLAYS