Evaluation of Bonded Concrete Overlays on Asphalt Pavements (2022)

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.

5   State highway and local agencies have been introduced to BCOAs over the past two decades, with most BCOA projects occurring in the Midwest. As with the adoption of any new pavement technology, an evolutionary process is needed to construct projects and observe the resulting performance, to develop design tools and performance models, and to compile best practices and communicate those to pavement engineers. Given that a substantial number of BCOA projects are now more than 10 years old and that several key performance studies have been completed, wider application around the United States is possible. To track concrete overlay projects in the United States, the American Concrete Pavement Association (ACPA) created the National Concrete Overlay Explorer application, an open online resource for viewing a list of concrete overlays constructed in the United States from the early 1900s to today. As of 2016, 106 BCOAs were listed in Overlay Explorer (ACPA 2020). Overlay Explorer data come from voluntary uploads; therefore, the distribution of projects may not accurately represent agency use and experience with BCOAs. Almost all BCOA projects reported in Overlay Explorer were constructed in the past 25 years, and most projects have less than 20 years of service. The majority of projects were recorded as street or road concrete overlays with 30 state highway projects. A few practices used in the first concrete overlays from 20 years ago would not be considered best practices today. Another interesting finding from Overlay Explorer is that 48 of 108 reported BCOA projects (or 44%) used fiber-reinforced concrete (FRC). This is one key example of how BCOA practice has evolved, as early projects excluded the use of FRC; however, FRC is used extensively in BCOA projects today. Past Performance Studies Several studies have been conducted over the last four decades to document BCOA perfor- mance. The following describes previous studies and briefly summarizes their findings. • NCHRP Synthesis 99: Resurfacing with Portland Cement Concrete (Hutchinson 1982). This synthesis documented the history of concrete resurfacing from the start of the 20th cen- tury until the early 1980s. Five case studies of bonded concrete overlays of composite pave- ment were reported. These five studies, constructed in the 1940s, can be considered the first attempts at bonding concrete to an underlying asphalt layer. These early bonded concrete overlays of composite pavements had relatively thin concrete sections, a reflection of the traffic demands of the day. • NCHRP Synthesis 204: Portland Cement Concrete Resurfacing (McGhee 1994). NCHRP Synthesis 204, an extension of the work conducted for NCHRP Synthesis 99, discussed the first two known BCOAs. The synthesis described the 1991 Louisville “ultrathin whitetopping” C H A P T E R 2 Literature Review

6 Evaluation of Bonded Concrete Overlays on Asphalt Pavements experiment (Cole 1992; Mack, Cole, and Mohsen 1993), which many consider the first major experimental BCOA test section. This synthesis also highlighted the use of FRC for concrete overlays, even on thinner (2- to 4-in.) bonded concrete overlays. Both bonded and unbonded concrete overlay performance were enhanced using FRC. Many projects cited in the report used steel fiber–reinforced concrete; however, polypropylene fibers were also mentioned. The polypropylene fibers were likely the low-modulus synthetic fibers introduced primarily to control plastic-shrinkage cracking. McGhee (1994) reported that some slabs did not achieve full bonding, but the thinner concrete slabs performed better than anticipated. The synthesis also noted that, despite many of the concrete overlays being constructed as unbonded overlays over asphalt base, they still developed some degree of bonding, adding to the pavement system’s structural capacity. • Design and Concrete Material Requirements for Ultra-Thin Whitetopping (Roesler et al. 2008). This study resulted in the development of a state-of-the-art BCOA design method for the Illinois DOT based on the ACPA BCOA Thickness Designer method. The study also reviewed and performed laboratory tests on appropriate materials for concrete overlays. The main contributions of the research project were an Illinois DOT BCOA thickness design method, a potential performance enhancement attributable to the use of synthetic macro- fibers, and a testing procedure to characterize FRC as an input for the structural design method. The developed design method also includes characterization of FRC containing syn- thetic macrofibers, which directly influences the thickness design of the concrete overlay. Residual strength was used as a structural input for the design on the basis of ASTM C1609 test results. Subsequent design methods have used the Illinois DOT approach for character- izing FRC design (Harrington and Fick 2014). • Development of Design Guide for Thin and Ultra-Thin Concrete Overlays of Existing Asphalt Pavements. Task 1 Report: Compilation and Review of Existing Performance Data and Information (Vandenbossche et al. 2011). This paper reviewed the performance of 51 BCOA projects (including field applications, traffic studies, and accelerated pavement testing) across 11 states (Illinois, Iowa, Michigan, Minnesota, Mississippi, Missouri, New York, Oklahoma, Pennsylvania, Texas, and Virginia). The review also included the FHWA accelerated load facility sections constructed in the late 1990s at the Turner–Fairbank Highway Research Center. The following findings were reported: – BCOA performance largely depends on support conditions. Performance on US-169 in Minnesota and at the MnROAD testing facility indicated the minimum asphalt thickness is not to be less than 3 in., and the asphalt layer is to be in fair to good condition. – Corner cracking is the primary BCOA distress type when traffic is channelized at or near the longitudinal joint (e.g., 4- × 4-ft slab size). Corner cracking occurs because of rutting in the support layers and debonding of the concrete and asphalt layers. Other observed dis- tress types included transverse and longitudinal cracking, reflection cracking, joint faulting, joint spalling, buckled slabs (blowups), and popouts. – Use of dowel bars at transverse joints improves the performance of thicker BCOAs under high traffic (i.e., on interstates), but dowel bars are typically not necessary. Joint sealing can also help extend the performance life of BCOAs. – The potential for reflection cracking will increase if the flexural stiffness of the concrete layer is less than the flexural stiffness of the asphalt layer. Flexural stiffness is defined as D Eh( )= − µ12 1 (1) 3 2 where E = Elastic modulus (psi, i.e., pounds per square inch), h = layer thickness (in.), and µ = Poisson’s ratio.

Literature Review 7   – Movement of heavy truck loads increases tensile stress at the bottom of the concrete overlay, accelerating crack propagation. – Milling of the asphalt layer before placement of the BCOA layer is one of the best proce- dures for ensuring the best possible bond. – When longitudinal joints coincide with the wheelpath, the BCOA is susceptible to corner cracking. • Performance History of Concrete Overlays in the United States (Fick and Harrington 2014). This technical brief broadly describes both the history of concrete overlay construction in the United States and all types of concrete overlays. Two 2001 BCOA case studies were summa- rized. The first, a 1.5-mi project in Oklahoma on US-69, included placement of a 6-in. BCOA on the truck lane (7- × 6-ft slab size) and a 4-in. BCOA on the passing lane (6- × 6-ft slab size). The Oklahoma project also included synthetic macrofibers (3 lb/yd3) and skewed transverse joints. The second, on SR-16 in Montana, used a 4-in. BCOA placed on a cold-milled asphalt pavement. This project also reported using synthetic macrofibers (3 lb/yd3), variable asphalt thickness, and an approximate slab size of 4- × 4-ft. At the end of 13 years, the Oklahoma project supported 10.1 million equivalent single axle load (ESAL) repetitions, and the Montana proj- ect supported 2.2 million ESALs. Furthermore, the BCOA site investigations showed little distress (<1% cracked slabs) on either project (Figure 1 and Figure 2). Figure 2. BCOA performance on Montana SR-16, showing area repaired after construction issues marked in green. (Source: Fick and Harrington 2014.) Figure 1. BCOA performance on Oklahoma US-69. (Source: Fick and Harrington 2014.)

8 Evaluation of Bonded Concrete Overlays on Asphalt Pavements • Guide to Concrete Overlays: Sustainable Solutions for Resurfacing and Rehabilitating Exist- ing Pavements (Harrington and Fick 2014). Now in its third edition, this guide has become a well-known reference on the use of concrete overlays, including BCOAs. The guide provides an overview of overlay-type selection, appropriate surface preparation, material selection criteria, design guidance, construction practices, and site management. The guide also covers miscellaneous areas of design such as curb and gutter milling and resurfacing, doweling techniques, accelerated construction practices, and traffic control. Appendix C of the guide discusses the use of synthetic macrofibers, and Appendix D presents the application of 3-D laser- scanning technology for concrete overlay placement. • Structural Performance of Ultra-Thin Whitetopping on Illinois Roadways and Parking Lots (King and Roesler 2014a). This performance review of BCOA projects in Illinois sought to determine whether the design method proposed by Roesler et al. (2008), using synthetic macrofibers, was performing well or needed adjustments. King and Roesler (2014a) evalu- ated 20 concrete overlay projects, noting that most were performing well given the condition of the asphalt, current traffic level, and expected design life. Several BCOA sections built in the late 1990s were severely distressed, however, they had design features no longer recom- mended. For example, BCOAs featuring 4- × 4-ft slab sizes that did not use FRC experienced slab migration and misalignment, especially at intersections and when subject to heavy truck traffic (Figure 3). Falling weight deflectometer (FWD) tests were conducted on multiple Illinois projects. To effectively use deflection basin data, this study developed a new BCOA backcalculation method to account for the small slab size, a condition violating the traditional backcalculation assump- tions. In a separate study, King and Roesler (2014b) proposed using the back calculated effective concrete layer thickness to assess the structural capacity and bond between the concrete and asphalt layers. LTE across the short-jointed slab was also characterized and found to be high. • MnDOT Thin Whitetopping Selection Procedures (Taylor et al. 2017). This report describes a process to both qualitatively and quantitatively assess the applicability of a BCOA solution, given existing asphalt surface distress, asphalt layer thickness, and asphalt layer quality (e.g., age). Main distress types to consider included asphalt surface rutting, asphalt layer strip- ping, and fatigue cracking. The proposed selection procedure incorporated coring, preferably before a visual inspection, to quantify whether an existing pavement was a likely candidate for a BCOA. The assessment process had four general selection criteria: 1. A desk review to determine project limits, existing pavement structure, pavement condi- tion (current and over time), and traffic data. Figure 3. Slab migration in 4- ë 4-ft BCOAs that did not use FRC. (Source: King and Roesler 2014a.)

Literature Review 9   2. Coring to verify the quantity (a minimum of 3 in.) and condition (e.g., stripping, delamina- tion) of asphalt before placing a concrete layer. 3. A site visit to verify existing distress (Table 1), subgrade support, drainage conditions, profile grade and cross slope, vertical constraints, shoulders, and widened lanes. 4. Preliminary (optional) and final recommendations. • Concrete Overlay Performance on Iowa’s Roadways (Gross et al. 2017). In this study, BCOAs and unbonded concrete overlays on asphalt (UBCOAs) were found to perform better than bonded and unbonded concrete overlays on concrete. The majority (84%) of BCOA pave- ments in Iowa were in good to excellent condition. With respect to Pavement Condition Index (PCI), a higher percentage of BCOAs were in good to excellent condition compared with bonded concrete overlays on concrete (BCOCs) (Table 2). In relation to the International Roughness Index (IRI), the majority of data for all concrete overlays indicated acceptable to good condition. In addition, most BCOA projects had fewer than 40 faulted joints per mile, suggesting that faulting is not perceived as a problem with BCOAs in Iowa. • Guide for Concrete Pavement Distress Assessments and Solutions (Harrington et al. 2018). This guide summarized major distress types, causes, and solutions observed in concrete pave- ments. For BCOAs, distress types and causes are summarized in Table 3. The guide also discussed the mechanisms of distress development and the mitigation of its formation and growth through design, construction, material selection, and preventative maintenance tech- niques (Table 4). Finally, the guide described BCOA repair techniques, the most common being removal and replacement of distressed slabs. • Development of Performance Curves for Whitetopping in Minnesota (Burnham et al. 2019). This report summarized predictive performance models on the basis of the measured pave- ment condition of Minnesota whitetopping projects. The performance of 26 whitetopping projects and 21 past and current test sections at the MnROAD facility were included in the Distress Functional Classification Interstate Primary Secondary Fatigue cracking, wheelpath (%) 10 20 20 Longitudinal cracking, wheelpath (ft/mi) 550 1,250 1,250 Transverse crack spacing (ft) 130 50 50 Mean rut depth, both wheelpaths (in.) 2 2 2 Shoving, wheelpath (%) 4 15 30 Table 1. Recommended maximum distress levels in existing asphalt pavement. (Source: Adapted from Taylor et al. 2017.) Overlay Type (% of Data) Condition Category BCOC UBCOC BCOA UBCOA PCI 81–100 Excellent 21 51 54 58 61–80 Good 51 39 34 36 41–60 Fair 21 6 9 6 21–40 Poor 7 4 3 0 0–20 Very poor 0 0 0 0 IRI (in./mi) <95 Good 40 28 28 40 96–169 Acceptable 50 65 66 53 >170 Not acceptable 10 7 6 7 NOTE: BCOC = bonded concrete overlays on concrete; UBCOC = unbonded concrete overlays on concrete. Table 2. Iowa concrete overlay performance. (Source: Adapted from Gross et al. 2017.)

10 Evaluation of Bonded Concrete Overlays on Asphalt Pavements evaluation. Project characteristics included traffic loadings from lower-volume interstates to low-volume county roads, whitetopping thickness ranged from 4 to 8 in. (5 and 6 in. being the most common), panel sizes ranged from 6- × 6-ft to 15-ft long by 12-ft wide, remaining underlying asphalt thickness ranged from 3 to 14 in., and approximately half the projects included sealed joints. Observations indicated good or better performance (i.e., minimal cracking, faulting, and good ride quality) with little to no required maintenance. Predominant distress included longitudinal cracking, with limited transverse cracking and appreciable faulting on several projects. Core samples indicated a good bond between the concrete and asphalt layers through 5 years of service, after which the bond degraded (initiating near the transverse joints). Significant Findings from Past Pavement Studies As described, multiple studies have been conducted over the last several decades to review BCOA performance. A compilation of significant findings from these studies follows: • Synthetic macrofibers. Large-scale FRC testing of slabs under monotonic loading conditions demonstrated that synthetic macrofibers enhance the flexural capacity of concrete slabs (Roesler et al. 2004, 2006). These results originally led Illinois to apply synthetic macro fibers to BCOAs in the early 2000s. Most early BCOA reports and papers reported limited performance benefits for FRC; however, many of these early projects used microfibers or low-modulus macrofibers, Distress Type Cause Corner breaks Physical Inadequate slab thickness Load placement (longitudinal joints near wheelpath) Loss of bond between concrete and asphalt layers Inadequate slab edge support Materials/ chemical Asphalt layer stripping or raveling Concrete curl/warp Transverse cracking Physical Inadequate slab thickness Improper joint layout and resulting slab dimensions Late joint sawing Inadequate asphalt layer thickness or condition Inadequate preoverlay repair Longitudinal cracking Physical Inadequate slab thickness Improper joint layout and resulting slab dimensions Late joint sawing Inadequate asphalt layer thickness or condition Inadequate preoverlay repair Loss of bond between concrete and asphalt layers Materials/ chemical Nonuniform overlay support (asphalt stripping or raveling) Nonuniform overlay support: curl/warp of portland cement concrete Reflection cracking Physical Full-depth working crack in asphalt layer Layers well bonded; asphalt stiffness > concrete stiffness Wide transverse joints Physical Seasonal opening and closing of joints Incompressibles in poorly sealed and unsealed joints Transverse joint faulting Physical Migration of fines under traffic in presence of moisture Shifting of slabs under load on viscoelastic asphalt Inadequate load transfer Lane-shoulder joint spalling Physical Difference in seasonal movement Inadequate ties between travel lane and shoulder Compression failure at transverse joint Physical Incompressibles in joint Transverse overlay joints fail to activate properly Overlay construction under conditions allowing large joint movements and infilling of joints with incompressibles Table 3. Types and causes of BCOA distress. (Source: Adapted from Harrington et al. 2018.)

Literature Review 11   often at low quantities, not the high-modulus synthetic macro fibers and quantities currently in use today. Because of their structural benefit, synthetic macrofibers have primarily been used in BCOAs to reduce the required concrete slab thickness, especially on roadways with elevation con- straints (Roesler et al. 2008; Harrington and Fick 2014; King and Roesler 2014a; Gross et al. 2017). Synthetic macrofibers have also helped to minimize slab migration and misalignment and extend service life related to cracking (Roesler, Cervantes, and Amirkhanian 2012; King and Roesler 2014a; Gross et al. 2017). The concrete overlay guides and several design methods (e.g., Illinois DOT and ACPA BCOA) include discussion and application of synthetic macro- fibers (King and Roesler 2014a; Harrington et al. 2018). Distress Phase Mitigation Procedures Corner breaks Design Provide adequate overlay thickness Avoid placement of longitudinal joints within wheelpaths Material selection Ensure no stripping or raveling in asphalt surface layer Consider use of structural fibers Construction Avoid longitudinal joints within wheelpaths Provide clean, aggressively textured asphalt surface Control moisture and temperature conditions during placement Cure concrete promptly and effectively Preventive maintenance Maintain joint sealant Transverse cracking Design Ensure asphalt thickness and condition Provide adequate overlay thickness Use slab aspect ratio <1.5 and as close to 1.0 as possible Material selection Ensure no stripping or raveling in asphalt surface layer Consider use of structural fibers Construction Provide clean, aggressively textured asphalt surface Provide adequate preoverlay repairs Cure concrete promptly and effectively Ensure depth and timely sawcut of transverse joints Preventive maintenance Maintain joint sealant Longitudinal cracking Design Ensure asphalt thickness and condition Provide adequate overlay thickness Use slab aspect ratio <1.5 and as close to 1.0 as possible Material selection Ensure no stripping or raveling in asphalt surface layer Consider use of structural fibers Construction Provide clean, aggressively textured asphalt surface Provide adequate preoverlay repairs Avoid placing longitudinal joints in or near wheelpaths Cure concrete promptly and effectively Ensure depth and timely sawcut of transverse joints Preventive maintenance Maintain joint sealant Reflective cracking Design Increase concrete layer stiffness/thickness Decrease asphalt layer stiffness/thickness Material selection Use fiber or steel mesh reinforcement in concrete over crack Construction Perform full-depth repair of thermal crack before overlay placement Preventive maintenance None Wide joints/slab migration Design Provide lug anchors, longitudinal joint ties, or other restraints Material selection None Construction Keep pavement clean before joint sealing Preventive maintenance Maintain joint sealant Transverse joint faulting Design Provide dowel bars in thicker slabs Confirm adequate stability of asphalt layer Material selection Consider use of structural fibers Construction None Preventive maintenance Maintain joint sealant Longitudinal lane-shoulder joint spalling Design Use similar sections, shoulder ties, and drainage Material selection Use frost-resistant shoulder foundation materials Construction Protect the integrity of the subdrained system Preventive maintenance None Table 4. Types of BCOA distress, by project phase and associated mitigation procedures. (Source: Adapted from Harrington et al. 2018.)

12 Evaluation of Bonded Concrete Overlays on Asphalt Pavements Work at the University of Minnesota Duluth determined that synthetic macrofibers ini- tially improved LTE, differential displacement, and differential joint energy dissipation for specific fiber types and high dosages (Barman and Hansen 2018). The study also noted trade- offs for high fiber prices and issues with uniform mixing, such as balling of fibers. Long-term improvement in LTE attributable to synthetic fibers has not been observed. • Slab size. Early observations and research suggested slab size was an important factor in the performance of BCOAs (Mack, Cole, and Mohsen 1993; Risser et al. 1993). Slab size assisted in maintaining the bond between the slab and the existing asphalt layer, as well as in reduc- ing curling and load-induced stresses. Most BCOA design methods recommend using slab sizes between 4- × 4-ft and 6- × 6-ft while avoiding placement of longitudinal joints in the wheelpaths (although most corner breaks have occurred in 4- × 4-ft slabs) (Roesler et al. 2008; Vandenbossche et al. 2011; Li and Vandenbossche 2013). King and Roesler (2014a) suggest 6- × 6-ft slab sizes are preferred to minimize wheelpath channelization along the longitudinal contraction joint. • Concrete–asphalt interface and bonding. Construction activities to promote the bond between the concrete overlay and the underlying asphalt layer include roughening the sur- face by shot blasting, milling, high-pressure water blasting, and sand blasting (Harrington et al. 2018). Vandenbossche (2005) suggested that seasonal variations in asphalt layer resilient modulus affect the bond: in cold weather, slabs bonded to the asphalt would act together, but during warmer weather, the two layers could act as two separate entities, resulting in tension at the bottom of the concrete overlay. Research has observed the relative movement of the concrete and asphalt layers at various locations, especially at intersections or braking loca- tions (King and Roesler 2014a). Rasmussen and Rozycki (2004) reported that an ultrathin whitetopping project exhibited severe distress in sections paved on softer asphalt layers. Stripping, a major cause of premature interface deterioration, increases in moist environ- mental conditions. Stripping typically develops at the surface of the asphalt layer. To prevent debonding, the surface of the old asphalt layer can be cold milled to a depth below the layer susceptible to stripping. Cores may also be taken to check for stripping at depths below the depth of removal (Taylor et al. 2017). Mateos et al. (2017) described resistance to moisture damage as a primary criterion for BCOA design because moisture will permanently affect the asphalt layer stiffness. Mateos et al. (2015) similarly found that the concrete–asphalt bond improved with a stiff and fatigue- resistant asphalt mixture. Likewise, Vandenbossche et  al. (2011) showed that prolonged moisture exposure results in debonding of the interface and softening of the asphalt layer under heavy traffic. • Existing asphalt pavement condition and structure. Assessing asphalt pavement condition and thickness is essential before deciding on the concrete overlay type. Poor existing asphalt condition and inadequate thickness are the primary causes of early-age distress in concrete overlays. Case studies of early BCOA failures in Brazil, Taiwan, and other locations found multiple factors in common, including the poor condition of the existing asphalt layer and concrete materials with high cement content and low water-to-cementitious-materials (w/cm) ratios (Roesler et al. 2008). Depending on the distress and estimated future traffic level, an existing asphalt pavement showing substantial bottom-up fatigue cracking may not be an ideal candidate for a BCOA. Severe top-down fatigue cracking, as shown in over 10% of the wheelpaths on interstates and over 20% on primary and secondary roadways, needs to be removed by milling before BCOA placement (Taylor et al. 2017). Severe rutting and shoving (≥2 in.) in the asphalt layer also needs to be removed by milling. The milling operation contributes to the placement of a uniform concrete layer when conducted at the final profile. Ground-penetrating radar (GPR) and coring can be used to confirm that the recommended minimum 3-in. asphalt layer thickness remains after milling. King and Roesler (2014a) also

Literature Review 13   emphasized coring to confirm the uniformity of asphalt layer thickness and condition. The study referenced a project on Illinois SR-53, which showed heavy signs of distress in locations where the asphalt layer was thinner. A thin underlying asphalt layer, less than 3 in., is unlikely to effectively bond to and support a concrete overlay. Observations also showed thinner asphalt layers (<2.5 in.) can break under construction traffic and increase the chances for the over- lay joints to propagate through the full depth of the asphalt layer, significantly lowering LTE (Roesler et al. 2008). Existing transverse cracks in the asphalt layer can reflect through the BCOA if the calculated flexural stiffness ratio (as presented in Vandenbossche and Barman 2010) falls below 0.1. Taylor et al. (2017) suggested, where applicable on limited repair areas, applying a thin debonding strip (e.g., duct tape) over potential reflective cracks to mitigate the stress concentration. • Joint sealant. Vandenbossche et al. (2011) compared replicate BCOA sections (same traffic level) at MnROAD and found the sealed joint sections showed superior performance, specifi- cally reduced longitudinal cracking. Sealing joints is expected to improve BCOA performance, as water ingress contributes to interface delamination and asphalt stripping (Harrington and Fick 2014). However, joint seals have to be maintained. Partial sealing of joints often results in more degradation than if the joint were never sealed (Burnham 2014). In addition, slabs have buckled when incompressibles from gravel shoulders (sand, small stones, etc.) enter the unsealed joints (Burnham 2014). Reducing the number of joints by increasing the slab size from 4- × 4-ft to 6- × 6-ft may reduce water infiltration at the joints. King and Roesler (2014a) noted that contraction joints should be sawcut as narrow as possible to reduce joint-related distress (e.g., faulting, spalling). • Shoulder and edge support. Shoulder and edge support for a BCOA pavement are expected to improve long-term performance; however, no studies directly comparing the use of tied shoulders with other types of shoulders have been conducted on BCOA projects. In addition, transverse curling can be mitigated if the slab length (in feet) is 1 to 1.5 times the slab thick- ness (in inches). • FWD findings. FWD testing showed that most BCOA projects investigated in Illinois had high LTE (80% to 90%) when an asphalt layer was at least 3 in. thick. Transverse joints with high LTE (and an overlay from 1 to 6 years old) indicated either a nonactivated crack at the joints, particularly in slabs with shorter joint spacing, or excellent support and good slab contact (Roesler et al. 2008). • Doweling of BCOA pavements. In a project at MnROAD, doubling the joint spacing from 6- × 5-ft to 12- × 10-ft resulted in similar performance under the same traffic conditions (Vandenbossche et al. 2011). A main factor contributing to similar performance included placement of 1-in. diameter dowels along transverse joints in the 12- × 10-ft slabs. However, larger slab sizes have to be carefully considered, given their increased potential for mid-slab cracking caused by curling and loss in interface bond. Failure Modes The following conditions have been identified as major causes of poor BCOA performance (Taylor et al. 2017; Harrington et al. 2018): • Inadequate existing asphalt layer (in condition or thickness); • Poor design details, such as drainage, jointing, and cross slopes; • Poor BCOA construction practices; and • Concrete material–related distress. Harrington et al. (2018) recommended means for measuring and assigning severity levels to assess BCOA performance. Table 5 summarizes BCOA-related distress and measures to qualify severity.

14 Evaluation of Bonded Concrete Overlays on Asphalt Pavements Failure Mechanisms Rasmussen et al. (2002) described loading factors and conditions leading to failure in BCOAs under FHWA accelerated load facility testing. Transverse and longitudinal crack formation depended on wheelpath position and development of permanent deformation in the asphalt layer. Corner cracks were assumed to be the result of poor concrete-to-asphalt bonding. FHWA accelerated load facility project tests, with slab sizes ranging from 3- × 3-ft to 6- × 6-ft and thicknesses between 3.25 and 4.5 in., showed an overwhelming amount of corner breaks, some transverse and longitudinal cracking in 6- × 6-ft slabs, and minor longitudinal faulting. MnROAD also used ultrathin and thin whitetopping sections to investigate the concrete overlay failure mechanism. Factors largely influencing the mode of failure were a combination of slab geometry and wheelpath location at a longitudinal joint (Vandenbossche et al. 2011, Distress Description Measurement Severity Levels Corner breaks Full depth, intersects adjacent transverse and longitudinal joints at approximately 45 degrees, length of sides <½ slab width Number slabs at each severity level Repaired areas rated as a high-severity corner crack and as a patch Low: <1/16 in. wide, spalled <10% length, no broken pieces, no faulting Medium: 1/16–1/8 in. wide or spalled (low), >10% length, no broken pieces, faulting <0.5 in. High: >0.125 in. wide, >10% medium- or high-severity spalling >10% length, broken ≥2 pieces and may be loose, or faulting >0.5 in. Transverse and longitudinal cracking (full depth) Transverse: perpendicular to centerline Longitudinal: parallel to centerline or lane-shoulder joint Number slabs at each severity level Repaired areas rated as a high-severity crack and as a patch Low: <1/16 in. wide, spalled <10% length, no broken pieces, no faulting Medium: 1/16–1/8 in. wide or spalled (low) >10% of length, no broken pieces, faulting < 0.5 in. High: >1/8 in. or spalling (medium or high) >10% length, or ≥2 broken pieces, may be loose, or faulting >0.5 in. Reflective cracking Full-depth crack of the concrete overlay directly above an existing crack in the underlying asphalt pavement Number and length at each severity level Total length of the crack rated at highest severity level present (≥10% crack length) Low: <1/16 in. wide, spalled <10% length, no broken pieces, no faulting Medium: 1/16–1/8 in. wide or spalled (low) >10% length, no broken pieces, faulting <0.5 in. High: >1/8 in. or spalling (medium to high ) >10% length or ≥2 broken pieces, may be loose, or faulting >0.5 in. Transverse joint faulting Difference in elevation across a joint Measure using faultmeter in the outer wheelpath and report to nearest 0.04 in. Not applicable: Severity levels could be defined by categorizing measurements, but a complete record of measurements is generally more desirable Transverse joints/slab migration Joints are wider than constructed because of independent longitudinal movement (migration) of the slabs Measure width at top of joint sealant Measure nonspalled areas in outer wheelpath (to nearest 1/16 in.) Measure displacements with adjacent feature Not applicable: Severity levels could be defined by categorizing measurements, but a complete record of measurements is generally more desirable Longitudinal lane-shoulder joint spalling (caused by heave) Cracking, breaking, chipping, or fraying of slab edge Length of longitudinal joint affected at each severity level Record spalls ≥4 in. Rate repair areas as patches Low: <2 in. wide (measured to joint face), no material loss, no patching, or spalled with no material loss and no patching Medium: 2–4 in. wide (measured to the face of the joint) with material loss and no patching High: >4 in. wide or spalls containing patch material Table 5. Types and measurement of BCOA-specific distress. (Source: Adapted from Harrington et al. 2018.)

Literature Review 15   Vandenbossche and Sachs 2013a). The 4- × 4-ft slabs showed increased risk of distress at the corner, similar to the observation by Roesler et al. (2008) (Figure 4). Studies have shown that loading along the longitudinal joint in conjunction with wet weather conditions or saturated joints contributes to debonding of the concrete–asphalt interface, making the BCOA more susceptible to corner breaks (Vandenbossche et al. 2011). Vandenbossche and Fagerness (2002) noted that underlying thermal cracks in the asphalt layer and traffic and envi- ronmental loadings combined to produce cracks in BCOAs (Figure 5). Although shorter slab sizes (4- × 4-ft) were expected to mitigate this issue, a MnROAD study reported that larger slabs exhibited fewer reflective cracks (Vandenbossche et al. 2011). As indicated by Burnham et al. 4 ft x 4 ft panels Dashed Lines Indicate Location of Wheelpath. Figure 4. Corner breaks in 4- ë 4-ft BCOA slabs. (Source: Vandenbossche et al. 2011.) Reflection transverse cracks Figure 5. Reflection cracking in BCOA slabs. (Source: Vandenbossche et al. 2011.)

16 Evaluation of Bonded Concrete Overlays on Asphalt Pavements (2019) in an evaluation of BCOA projects in the Minnesota highway network, transverse reflec- tive cracking was actually more limited than originally thought. The first BCOA pavement in Illinois, located at the intersection of US-36 and Oakland Avenue in Decatur, had up to 6 in. of slab migration after 14 years of service (King and Roesler 2014a). Slab migration and misalignment of the 3- × 3-ft and 4- × 4-ft slabs were the result of decelerat- ing and accelerating heavy traffic (average annual daily traffic, or AADT, of approximately 17,500). Significant faulting at both the longitudinal and transverse joints was reported where truck wheelpaths coincided with the longitudinal joints. After 14 years, 36% of the slabs exhib- ited longitudinal and corner cracks. King and Roesler (2014a) observed that constructing BCOA with FRC containing synthetic macrofibers has mostly eliminated slab migration and misalignment. On MnROAD sections, significant distress included joint faulting, especially for large slab sizes and high traffic levels (Vandenbossche et al. 2011). Similarly, noticeable faulting occurred on a project in Illinois on Sailor Springs Road in Clay County. This project, a 5- to 6-in. thick BCOA constructed in 1998 and evaluated in 2012, exhibited no cracking but did show notice- able faulting on sections with 15-ft joint spacing and decreasing faulting on sections with 11-ft and 6-ft joint spacing (King and Roesler 2014a; IGGA 2017). Network-Level Performance Gross et  al. (2017) surveyed more than 380 overlay projects in Iowa, of which 178 were BCOAs. IRI and PCI were used to compare the network-level performance of the various over- lay options. Using a PCI threshold of 60 as the demarcation from fair to poor condition, the survey found that the trendline crossed this threshold at an average age of 25 years, 5 years beyond the expected design life of 20 years. The PCI threshold of 60 was reached at 35 years for 5-in. BCOAs and at 37 years for 6-in. BCOAs. Similarly, an IRI threshold of 170 in./mi was used to demarcate the transition from fair to poor pavement condition. The IRI trend line indicated the average service lives of the 4-, 5-, and 6-in. BCOAs were 35, 35, and 37 years, respectively (R2 = 0.56 to 0.30). However, the majority of the 4- to 6-in. BCOA projects were less than 10 years old, whereas thicker BCOA and unbonded concrete on concrete overlay projects were more than 20 years old. Burnham et al. (2019) summarized the performance of 26 whitetopping projects throughout Minnesota. Study findings indicated that Minnesota whitetopping projects (most being less than 10 years old) are performing well, with the majority receiving minimal maintenance. Pri- mary distress included longitudinal cracking caused by insufficient overlay thickness and lack of dowels, rare occurrences of transverse cracking caused by reflective cracking, and slab buckling caused by filling of joints with incompressibles on roadways with gravel shoulders. This study also developed a whitetopping performance prediction model based only on IRI. Preliminary study findings indicated a minimum terminal service life (IRI > 170 in./mi) of approximately 18 years, with some projects projected to reach 25 years of service. Concrete Materials Concrete mixture proportions have varied widely across BCOA projects. Earlier projects were designed to have rapid construction and early-opening-to-traffic requirements. This led to mix- ture designs with high cement contents and notable failures on some projects (Roesler et al. 2008). In more recent years, there has been a shift to more conventional mixtures but with the

Literature Review 17   addition of synthetic macrofibers. The following summarizes aggregates, cement content, and synthetic macrofiber use as related to mix design for BCOAs. Aggregate Typically, agencies specify that aggregates conform to ASTM C33 and be well graded (Harrington and Fick 2014). Various BCOA projects have performed successfully by using a variety of aggregate types. Roesler et al. (2008) studied the impact of mixture proportions and constituents on the 28-day fracture properties of BCOA mixtures by using the single-edge notched beam [SEN(B)] test specimen (Figure 6). The total fracture energy did not capture variations in mixtures with different cement contents or aggregate proportioning. However, aggregate type did change the total fracture energy, with higher total fracture energy in coarse river gravel aggregates than in crushed dolomite aggregates. Crushed dolomite aggregate con- taining 50% recycled coarse aggregate showed similar measured fracture energy relative to the control mixture. Maximum coarse aggregate size is a function of overlay thickness. Common maximum aggre- gate size has ranged from 0.75 to 1.0 in.; however, the largest practical maximum aggregate size can be selected to minimize paste requirements, reduce shrinkage, minimize cost, and improve aggregate interlock at joints and cracks (Harrington and Fick 2014). Cement Content Type I and II cements are commonly used in concrete overlays (Harrington and Fick 2014). Care needs to be taken with the design of high early-strength mixtures. Roesler et al. (2008) identified that multiple BCOA projects constructed in the United States and internationally developed premature cracking and failure related to excessive shrinkage. These BCOA projects had an objective to achieve high early strength, which drove the projects to use high cementi- tious content mixtures (Roesler et al. 2008). High cementitious content mixtures increase the material’s brittleness, increase drying and thermal shrinkage, increase curling potential of the slab, and raise the likelihood of debonding the concrete–asphalt interface. The cementitious Figure 6. Single-edge notched beam during testing. (Source: Roesler et al. 2008.)

18 Evaluation of Bonded Concrete Overlays on Asphalt Pavements proportion of BCOA mixtures can be minimized to achieve the specified strength with a recom- mended w/cm ratio of 0.40 to 0.42. Synthetic Macrofibers The use of synthetic macrofibers has been shown to improve the performance of thin concrete overlays (Harrington and Fick 2014). In a study by Roesler et al. (2008), the addition of synthetic macrofibers significantly improved the fracture energy of plain concrete at all ages. Harrington and Fick (2014) noted that synthetic macrofibers improved the performance of thin concrete overlays in four ways: 1. Increased concrete toughness, allowing thinner concrete slabs. 2. Minimized differential slab movement, allowing longer joint spacing. 3. Increased resistance to plastic shrinkage, improving performance. 4. Held cracks tightly together, improving performance. Synthetic macrofibers are commonly added at a rate of 3 to 4 lb/yd3, as specified in ASTM C1161, Type III, Section 4.1.3; however, ASTM C1609 provides an alternate method for deter- mining fiber content. Mix Design Table 6 summarizes the mix design components on several unbonded concrete overlay projects. Construction Practices Several key factors are required for the successful construction and long-term performance of BCOAs. First, the existing asphalt surface needs to be clean and any rutting (or shoving) greater than 2 in. needs to be removed by milling (Harrington and Fick 2014). Structurally distressed areas (e.g., potholes, areas of severe fatigue cracking) need to be removed and replaced before the overlay. Small quantities of transverse and longitudinal cracking can be ignored as long as the cracks are narrower than the maximum aggregate size of the concrete overlay (Harrington and Fick 2014). Severe transverse thermal cracks have been problematic for BCOA performance, and localized areas need to be repaired or covered with a bond breaker before the concrete over- lay. Finally, the remaining asphalt layer thickness is to be no less than 3 in. and in fair to good condition (Fick and Harrington 2015). Joint sealing (and resealing) has been shown to improve BCOA performance by limiting water ingress, especially at the longitudinal joints (Harrington and Fick 2014). Sealing can also minimize the amount of incompressibles entering the joints. Incompressibles have led to blowups on some BCOA pavement sections. Component (lb/cy) Average Quantity Coarse aggregate 1,850–1,900 Fine aggregate 1,000–1,250 Cement 550–750 Water 180–290 Table 6. Concrete overlay mix design quantities. (Source: Adapted from Mateos et al. 2015.)

Literature Review 19   The following step-by-step construction procedures are commonly advised in placement of BCOAs (Vandenbossche and Sachs 2013b; Fick and Harrington 2015): 1. Mill existing asphalt surface, if needed, to a depth to minimize debonding of underlying asphalt layer (at least 1 in. of lift to remain). Adjust changes in grade or cross slope by adjust- ing the concrete overlay thickness rather than by increasing milling depth. Set depth of milling to remove distortions greater than 2 in., match curb or existing structure elevations, and meet minimum vertical clearance requirements. 2. Clean asphalt surface by sweeping or using compressed air immediately before concrete placement to ensure proper bonding. 3. Mist asphalt surface to bring up to saturated, surface-dry condition and to minimize moisture being pulled from the concrete mixture (maintain asphalt surface temperature below 120°F). Do not allow water to puddle. 4. Place and finish concrete. Do not place concrete when air or existing pavement surface is below 40°F. If synthetic macrofibers are included, ensure they are uniformly distributed throughout the concrete mixture. 5. When required, place a maximum of No. 4 tie bars and dowel bars per agency requirements. 6. Apply liquid curing compound uniformly across the pavement surface and on all exposed edges. Ensure curing compound is not spilled on asphalt surface to be overlaid. 7. Saw joints to 1/3 the depth of the concrete layer and 1/8 to 1/4 in. wide. Backer rods are not required. Clean and seal joints with asphalt sealant as quickly as possible to prevent cracking. Crack sealing is recommended at locations with gravel shoulders to minimize incompress- ibles entering the joints and leading to buckling. 8. Open the road to traffic when recommended concrete strength is achieved, typically, 420 to 480 pounds per square inch (psi) flexural or 2,500 to 3,000 psi compressive. Transition from a concrete to an asphalt pavement can lead to performance issues if not properly designed. Figure 7 illustrates the Texas Department of Transportation standard plan to address this issue for thin whitetopping. Maintenance Practices Like all pavements, BCOAs inevitably develop distress. Common maintenance and reha- bilitation include removing and replacing distressed slabs and joint resealing (Vandenbossche and Sachs 2013b; Taylor et al. 2017; Harrington et al. 2018). Slab movements compromise the structural capacity of adjacent slabs if BCOAs are patched with asphalt. Summary Many studies have evaluated the performance of in-service BCOA projects and identified criteria essential for long-term performance. BCOAs rely on the underlying asphalt pavement to provide structural support. Feasible projects include asphalt pavements exhibiting rutting or shoving of less than 2 in., allow for at least 3 in. of asphalt pavement in fair to good condition to remain in place after milling, and are otherwise structurally sound. Typical slab sizes range from 4- × 4-ft to 12- × 12-ft with BCOA layer thickness of 3 to 7 in., although early BCOA projects were constructed using smaller slab sizes. Based primarily on MnROAD test sections, critical failure modes are corner breaks for 4- × 4-ft slab sizes, longitudinal cracking for 4.5- × 7-ft slab sizes, and transverse reflection cracking for 12-ft wide slabs. BCOA concrete mixture proportions have varied significantly among evaluated projects. Early projects focused on rapid construction and early opening to traffic, which resulted in high

Figure 7. Texas DOT standard plan for thin whitetopping at intersections. (Source: Texas DOT 2004.)

Literature Review 21   cement contents and early failures. More recently, BCOAs have used conventional concrete mixtures with or without synthetic macrofibers. BCOAs can be placed with conventional concrete paving equipment if considerable attention is paid to the condition of the asphalt layer, specifically after severe rutting and shoving greater than 2 in. is removed by milling and severely fatigued areas are removed and replaced. Whereas small quantities of transverse and longitudinal cracking can be ignored, transverse cracking wider than the maximum aggregate size of the BCOA needs to be repaired. Maintenance on BCOAs typically includes joint resealing, crack sealing, and removing and replacing distressed slabs. Key findings related to BCOA project selection, design, construction, performance, and main- tenance are summarized in Table 7. Table 7. Summary of BCOA project activities. Criteria Recommendation Selection of suitable candidates Stable structural support conditions Existing pavement condition Functional Classification Fatigue Cracking (%) Longitudinal Cracking (ft/lane-mi) Transverse Crack Spacing (ft) Wheelpath Shoving (%) Mean Rut Depth (in.) Interstate <10 <550 <130 <4 <2 Primary <20 <1,250 <50 <15 <2 Secondary <20 <1,250 <50 <30 <2 Minimum of 3 to 4 in. of asphalt pavement to remain after milling Good drainage conditions Predesign activities Review historical records Conduct site visits Conduct coring Optional FWD testing (backcalculate layer modulus) Optional GPR testing (layer thickness) Design considerations Avoid placing longitudinal joint in wheelpath; slab size >4- x 4-ft Consider dowel and tie bars (for larger panels) when BCOA ≥5 in. and heavy truck traffic Saw joints T/3 deep, seal with asphalt or other applicable sealant Consider adding synthetic macrofibers on the basis of truck loadings Concrete material properties Well-graded aggregate (ASTM C33), maximum aggregate size 0.75 to 1 in. Type I or II cement, with w/cm ratio of 0.40 to 0.42 Preoverlay repair Mill severe rutting and shoving ≥2 in. Clean and fill thermal cracking (crack width > max aggregate size) Repair potholes using a full-depth concrete patch Construction Mill to enhance bond, especially for BCOA <4 in. thick Clean surface (sweeper, compressed air) Mist surface, place, and finish concrete Apply curing compound Sawcut and seal joints Performance (predominant distress) <4.5-ft joint spacing: corner breaks 4.5-ft x 7-ft joint spacing: longitudinal cracking 12-ft wide slabs: conventional concrete distress Maintenance activities Reseal joints and seal cracks Localized panel replacements and patching (preferably with like material)