Guide to Using Existing Pavement in Place and Achieving Long Life (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.

169 INTRODUCTION Long-life pavement is defi ned in this document as pavement sections designed and built to last 50 years or longer without requiring major structural rehabilitation or reconstruction. Periodic surface renewal activities are expected over the 50-year dura- tion. Long-lasting concrete pavements are readily achievable, as evidenced by the num- ber of pavements that remain in service in excess of 50 years; however, recent advances in design, construction, and materials provide the knowledge and technology needed to consistently achieve this level of performance. A more detailed working defi nition as suggested by Tayabji and Lim (2007) of long-life concrete pavement includes the following: • Original concrete service life is 40+ years. • Pavement will not exhibit premature construction and materials-related distress. • Pavement will have reduced potential for cracking, faulting, and spalling. • Pavement will maintain desirable ride and surface-texture characteristics with minimal intervention activities, if warranted, for ride and texture, joint resealing, and minor repairs. • Life-cycle costs and user costs will be reduced. The pursuit of long-life concrete pavements requires an understanding of analy- sis, design, and construction factors that affect short- and long-term pavement perfor- mance. This requires an understanding of how concrete pavements deteriorate and fail. Photos of completed and under-construction jointed plain concrete pavements (JPCPs) and continuously reinforced concrete pavements (CRCPs) are shown in Figure 3.1. 3 RIGID PAVEMENT BEST PRACTICES

170 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Pavement Distress Thresholds Generally recognized threshold values in the United States for distresses at the end of the pavement’s service life are presented in Table 3.1 for JPCP and CRCP. These failure mechanisms can be addressed through application of best practices for structural design (layer thicknesses, panel dimensions, joint design, base selection, and drainage considerations), material selection (concrete ingredients, steel, and foun- dation), and construction activities (compaction, curing, saw-cut timing, surface tex- ture, and dowel alignment). The trends in structural design of rigid pavements have generally resulted in thicker slabs and shorter joint spacings (for JPCP) along with widespread use of corrosion-resistant dowel bars and stabilized base layers (especially asphalt-stabilized base layers). (a) (b) (c) (d) Figure 3.1. Completed and under-construction JPCP and CRCP. (a) and (b) JPCP constructed on HMA base. (c) and (d) CRCP constructed on HMA base. Photos: Joe Mahoney.

171 RIGID PAVEMENT BEST PRACTICES TABLE 3.1. THRESHOLD VALUES FOR CONCRETE PAVEMENT DISTRESSES Distress Threshold Value Cracked slabs, % of total slabs (JPCP) 10%–15% Faulting (JPCP) 0.25 in. Smoothness (IRI), m/km (in./mi) (JPCP and CRCP) 2.5–3.0 (150–180) Spalling (JPCP and CRCP) Minimal Material-related distress (JPCP and CRCP) None Punchouts, number/mi (CRCP) 12–16 Note: IRI = international roughness index. Source: Tayabji and Lim, 2007. Types of Concrete Overlays To design and construct long-lasting rigid pavement overlays as applied to existing pavements, it is important to define the three types of concrete overlays. Typical con- crete overlay types were described by Rasmussen and Rozycki (2004). Even though the industry has defined improved terminology and definitions for concrete overlays, these original terms are still widely used and are described below: • Unbonded concrete overlays consist of a portland concrete cement (PCC) layer constructed on top of an existing PCC pavement, separated by a bond breaker. • Bonded concrete overlays consist of a PCC layer constructed on top of an existing PCC pavement, bonded to the existing pavement. • Whitetopping involves a PCC layer constructed on top of an existing hot-mix as- phalt (HMA) pavement. Subcategories of whitetopping include thin whitetopping (TWT) and ultrathin whitetopping (UTW): — Conventional whitetopping overlays are ≥8 in. thick, — TWT overlays are >4 in. but <8 in. thick, and — UTW overlays are ≤4 in. thick. An illustration of the different types of concrete overlays is shown in Figure 3.2. Figure 3.2. Types of concrete overlays—earlier descriptions. Source: Rasmussen and Rozycki, 2004.

172 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE The Texas Department of Transportation (TxDOT) recommends a design life of only 5 to 10 years for bonded concrete overlays of asphalt pavements for a range of PCC overlay thicknesses from 4 to 7 in. (greater thickness is associated with higher truck traffic) (Texas Department of Transportation, 2011). Anecdotally, other states have reported using design lives of 20 years or more for similar bonded concrete over- lay designs. TxDOT uses the term “thin whitetopping” in its Pavement Design Guide (PDM) (Texas Department of Transportation, 2011) to describe this type of overlay, which is normally used at intersections where rutting and shoving of HMA causes performance problems. The TxDOT PDM notes that the contraction joints are to be spaced 6 ft apart with all panels being square. More recent concrete overlay terminology was described by Harrington (2008). The new definitions provide a simplified description of concrete overlays as shown in Figure 3.3. Two categories are shown: (1) unbonded concrete overlays and (2) bonded concrete overlays. Subcategories are defined based on the underlying pavement, which can be (1) concrete, (2) asphalt, or (3) composite pavements. Figure 3.3. Types of concrete overlays—more recent descriptions. Source: Harrington, 2008.

173 RIGID PAVEMENT BEST PRACTICES RIGID PAVEMENT RENEWAL STRATEGIES The renewal strategies for long life using existing pavements as described in this best practices chapter are • Unbonded concrete overlays of concrete pavements and • Unbonded concrete overlays of HMA pavements. The logic for selecting these two long-life strategies follows. SUPPORTING DATA AND PRACTICES Long-life renewal strategies should be designed as a system that covers a combination of materials, mixture and structural design, and construction activities. Smith, Yu, and Peshkin (2002) state that the success of long-life renewal alternatives using existing pavements hinges on two critical parameters: (1) the timing of the renewal and (2) the selection of the appropriate renewal strategy. The timing and selection of the appro- priate renewal strategy are dependent on factors such as the condition of the exist- ing pavement; the rate of deterioration of the distress; the desired performance life from the repair strategy; lane closures and traffic control considerations; and user costs. Given the definition of long-life renewal strategies and the constraints of life expectancy associated with timing and selection of pavement renewal strategies, only unbonded concrete overlays (using HMA separator layers) of existing concrete and asphalt pave- ments are likely to perform adequately for 50 or more years. This conclusion is based on several sets of information which includes, but is not limited to, (1) prior pavement design criteria, (2) state DOT criteria and field projects, (3) LTPP findings, (4) state field visits, and (5) information from the National Concrete Pavement Technology Center (Harrington, 2008). It is and has been apparent that slab thickness is a major factor in long-life renewal options. Well-known design procedures for PCC systems have been available for sev- eral decades. For example, Packard (1973) used fatigue concepts for airport pavement design for the Portland Cement Association (PCA). Packard (1973) and Neville (1975) both noted that for flexural stress ratios less than 0.55 (applied flexural stress divided by modulus of rupture), the fatigue life of PCC is unlimited. Packard actually used a stress ratio of 0.50 to add a bit of conservatism to the PCA airfield design process. Additionally, Packard (1984) produced a fatigue-based highway design method for PCA. This method is also based on fatigue principles [specifically, the flexural stress is divided by the modulus of rupture (28-day cure)]. These fatigue-based approaches use Miner’s hypothesis (Miner, 1945) for accumulating fatigue damage. In addition to existing design procedures and state DOT practices, an extensive amount of pavement performance data has been collected over the past 20 years via the Long-Term Pavement Performance (LTPP) program. These results, as relevant to long-life rigid pavement renewal best practices, are summarized as follows.

174 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Long-Term Pavement Performance (LTPP) and State DOT Information LTPP LTPP results were examined to see what could be learned about long-life designs. This included data from General Pavement Study 9 (GPS-9) and Special Pavement Study 7 (SPS-7) projects. Unbonded Concrete Overlays From the GPS-9 experiment (“Unbonded Concrete Overlays,” which included un- bonded JPCP or CRCP overlays placed on JPCP or CRCP), performance data re- viewed for Phase 1 of this study were used. The overlay thicknesses ranged from 5.8 to 10.5 in. Separator layers included dense-graded asphalt concrete, open-graded asphalt concrete, and chip seals. The average joint spacing was about 16 ft and load-transfer mechanisms were either aggregate interlock or steel dowels. A summary of the sections and major findings from that assessment include the following: • Of the unbonded overlays reviewed, the thicknesses were — ~6 in. thick, 22%; — ~8 in. thick, 22%; — ~9 in. thick, 11%; and — ∼10 in. thick, 45%. • The thicker JPCP overlays (≥8 in.) exhibited essentially no transverse cracks. The CRCP overlays had transverse cracks with ~4-ft spacing for overlays <10 in. thick and ~5-ft spacing for overlays >10 in. thick. • On average, thicker GPS-9 overlays had lower IRI values. • The overall magnitude of the faulting was well below 0.25 in. for all unbonded overlays (the threshold considered for long-life pavements). Faulting levels were significantly less for (1) thicker slabs (∼10 in. thick), (2) interlayer thicknesses >2 in., and (3) use of HMA as the interlayer material. • Thicker HMA interlayers appear to inhibit transverse cracking. This condition also contributed toward the integrity of the joint by controlling the amount of joint faulting. • Use of dowel bars in transverse joints had a positive impact on all pavement per- formance measures. Bonded Concrete Overlays From the SPS-7 experiment (“Bonded Concrete Overlays on PCC Pavement”), these sections were examined for Phase 1 of this study and included three types of bonded overlays: JPCP, CRCP, and plain concrete pavement (PCP). The third type of overlay included PCP, which was placed on existing CRCP but without reinforcement in the overlay. The ages of overlays ranged from 7 to 11 years (the time between construc- tion and the last condition survey). The overlay thicknesses of the various test sections ranged from a minimum of 3.1 in. to a maximum of 6.5 in. The bonding agent type

175 RIGID PAVEMENT BEST PRACTICES used in 21 of the SPS-7 sections was water with cement grout, and in 13 sections no bonding agents were employed. The surface-preparation methods used to create bond in the various sections included shot blasting, water blasting, and milling. The major findings from that assessment follow: • Of these overlays located in four states, the total number of sections (35) expressed as percentages associated by overlay type are — CRCP, 51%; — JPCP, 26%; and — PCP, 23%. • For bonded JPCP overlays, eight sections all were located on Route 67 in Missouri—which, at the time of construction (1990), experienced about 250,000 equivalent single axle loads (ESALs)/year. The JPCP overlays ranged in thickness from 3.0 to 5.4 in., with an average of 4.3 in. These overlays were placed on exist- ing JPCP, which had a 20-ft spacing between transverse joints. Before the bonded overlays were placed, two surface-preparation treatments were used: either shot blasting or milling. All of these SPS sections had a length of 500 ft. The actual overlay thicknesses and performance with respect to transverse cracks over 5 years following construction are shown in Table 3.2. TABLE 3.2. OVERLAY THICKNESS AND PERFORMANCE OVER 5 YEARS Target Overlay Thickness (in.) Overlay Thickness Based on Cores (in.) Number of Transverse Cracks Before Overlay (JCPC constructed in 1955, 10-in. slabs) Number of Transverse Cracks 5 Years After Construction 3.0 4.4 1 21 3.0 3.0 0 11 3.0 3.6 9 43 3.0 3.0 0 15 5.0 4.8 6 102 5.0 4.9 3 101 5.0 5.2 2 94 5.0 5.4 4 130 Source: Smith and Tayabji, 1998; Missouri DOT, 1998. • Given the cracking levels observed for these nominal 3- and 5-in.-thick bonded overlays, it is unlikely these sections will serve adequately for 50 years. The Missouri DOT notes the following in its Missouri Guide for Pavement Rehabilita- tion (2002): “(1) A bonded PCC overlay is a viable rehabilitation treatment that has historically been technically difficult to construct properly, and (2) unbonded PCC overlays should provide at least 20 years of good performance if properly de- signed and constructed. PCC thickness should be ≥8 inches with an AC interlayer

176 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE ≥1 inch.” Thus, use of bonded overlays is allowed but unbonded overlays are preferred with 8-in. or thicker slabs. • The CRCP overlays ranged in thickness from 3.2 to 6.5 in. with an average of 4.6 in. All of these overlays were placed on existing CRCP. • The CRCP overlays show more promise in that only 4 of 19 sections in the SPS-7 experiment exhibited punchouts following 5 to 7 years of service; however, the length of service precludes a clear view about longevity. • The data suggest that, on average, thicker SPS-7 overlays (>6 in.) resulted in lower IRI values. Given the performance of the LTPP JPCP bonded concrete overlays in Missouri and the amount of cracking observed, this study and the rigid best practices will focus only on unbonded concrete overlays over existing concrete and flexible pave- ments. The amount of transverse cracking suggests that a 50-year life is only likely for unbonded overlays. This is further supported by additional state experience, which follows. The exception might be bonded CRCP overlays, but additional performance data are desirable. Texas Department of Transportation CRCP Overlays During the conduct of the SHRP 2 R23 study, a field trip to review concrete overlays was made with TxDOT. Most of TxDOT’s bonded concrete overlays are located in the Houston area and are CRCP overlays over existing CRCP. Based on observed perfor- mance of 4- to 8-in.-thick bonded overlays and views expressed by TxDOT personnel, it appears that bonded CRCP overlays within that thickness range can be expected to perform for about 25 years. One unbonded 12-in.-thick CRCP overlay approximately 10 years old at the time of visit was performing well. Information by Kim et al. (2007) documented the performance of 4-in. bonded concrete overlays on existing CRCP in Houston on I-610. The 4-in. overlays were reinforced with either wire mesh or steel fibers. The existing CRCP was assessed to be structurally deficient with 8 in. of CRCP over 1 in. of HMA over 6 in. of CTB. After 20 years of service, the wire-mesh overlay sections provided the best performance in the experiment along with the use of limestone aggregate [a material with low coefficient of thermal expansion (CTE)]. Washington State DOT Bonded Concrete Overlays Bonded JPCP concrete overlays constructed in 2003 over existing HMA were re- viewed (Figure 3.4). Three thicknesses of concrete overlays were used: 3, 4, and 5 in., each placed on I-90 east of Spokane, Washington, which experiences about 1,000,000 ESALs/year. These sections were removed during 2011 due to pavement reconstruction; thus, they were in service for 8 years. Each of the bonded concrete overlays was 500 ft long and used the same PCC mix. Transverse contraction joints were sawed at 5-ft spacings and the longitudinal joint split the 12-ft-wide lane (thus a joint spacing of 5 ft by 6 ft), as illustrated in Figure 3.5. The mix had a specified minimum flexural strength of 800 psi with a minimum cement content of 800 lb/yd3. Polypropylene fibers were added at a rate of 3 lb/yd3. A carpet

177 RIGID PAVEMENT BEST PRACTICES drag finish was applied to the surface (Anderson et al., 2006). The underlying HMA thicknesses were 9 in. for the 3-in. slab, 8 in. for the 4-in. slab, and 7 in. for the 5-in. slab. Following 1 year of service, cracking in the three bonded JPCP sections were as follows: • 87% of the 3-in.-thick panels were cracked, and • Each of 4- and 5-in. sections had 4% cracked panels. At the time of removal in 2011 (Figure 3.6), the 3-in. section was severely dis- tressed, as shown in Figure 3.5. The 4- and 5-in.-thick sections were in substantially better condition. The total accumulated ESALs at the time of removal were a bit less than 10 million. Figure 3.4. Construction of bonded PCC overlays that were placed directly on rotomilled HMA (Washington State, July 2003). Photos: WSDOT. Figure 3.5. Condition of 3-in. bonded PCC overlay of HMA in 2011, following 8 years of service. Photos: WSDOT.

178 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Minnesota DOT and MnRoad Bonded Concrete Overlays The Minnesota DOT constructed its first set of bonded JPCP concrete overlays on existing HMA at MnRoad in 1997, and it included 3-, 4-, and 6-in.-thick sections. Following 7 years of service, the 3- and 4-in.-thick sections were removed (Burnham, 2008). The 6-in. sections remained in service through 2010. Figure 3.7 shows the 3-in.-thick sections with two different joint layouts. The conclusion was the 5 ft by 6 ft joint layout was superior to the 4 ft by 4 ft layout, but the amount of cracking for both configurations was extensive. Figure 3.6. Removal of 3-in. PCC overlay before reconstruction of this portion of I-90. Bond between the PCC overlays was assessed visually during removal in 2011. Photo: WSDOT. (a) (b) Figure 3.7. Condition of 3-in. bonded concrete overlays following 5 million ESALs and 6 years of service. (a) MnRoad Cell 95. Bonded concrete overlay 3 in. thick with a 5 ft by 6 ft joint spacing in November 2003. (b) MnRoad Cell 94. Bonded concrete overlay 3 in. thick with a 4 ft by 4 ft joint spacing in November 2003. Photos: MnDOT.

179 RIGID PAVEMENT BEST PRACTICES Table 3.3 contains a summary of the 3-, 4-, and 6-in. sections. The applied ESALs are about 1,000,000/year on this portion of I-94. The 6-in. sections have survived through 2010 achieving an age of ≥13 years. Figure 3.8 illustrates the performance of the 6-in. sections at MnRoad following 11 years of service. TABLE 3.3. INITIALLY CONSTRUCTED MNROAD BONDED CONCRETE OVERLAY SECTIONS Cell Type PCC Thickness (in.) HMA Thickness (in.) Panel Size (ft) Year Start to Year End 92 TWT 6 7 10 × 12 (doweled) 1997–2010 93 UTW 4 9 4 × 4 1997–2004 94 UTW 3 10 4 × 4 1997–2004 95 UTW 3 10 5 × 6 1997–2004 96 TWT 6 7 5 × 6 1997 to present 97 TWT 6 7 10 × 12 1997–2010 Source: After Burnham, 2008. Recap on Concrete Overlays There are two types of bonded concrete overlays for which state and LTPP perfor- mance data are available: • Bonded JPCP concrete overlays over HMA and • Bonded concrete overlays over existing PCC. Given the information summarized, the performance of bonded JPCP concrete overlays over existing HMA is a function of slab thickness and design details such as joints and remaining HMA thickness. Given Interstate types of traffic (∼1 million ESALs per year), Table 3.4 shows typical pavement lives that can be expected for vari- ous slab thicknesses along with joint details. The expected lives shown are tentative and reflect an extrapolation of the field data reviewed. TABLE 3.4. BONDED CONCRETE OVERLAYS OVER EXISTING HMA WITH 1 MILLION ESALS PER YEAR WITH SUFFICIENT EXISTING HMA THICKNESS Slab Thickness (in.) Joints Dowels? Expected Life (years) 3 5 ft by 6 ft No 5 4 5 ft by 6 ft No 5–10 5 5 ft by 6 ft No 10–15 6 5 ft by 6 ft No 15–20 Note: It is assumed for all HMA thicknesses that the existing HMA materials are in good condition and exhibit no stripping.

180 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE (a) (b) (c) (d) Figure 3.8. Condition of 6-in. bonded concrete overlays following 10 million ESALs and 11 years of service at the time of the photos (Constructed in 1997). (a) MnRoad Cell 96. Bonded concrete overlay 6 in. thick with a 5 ft by 6 ft joint spacing without dowels. Performance: No cracked panels but noticeable faulting has occurred. Will be diamond ground in 2011 to improve ride. (b) MnRoad Cell 97. Bonded concrete overlay 6 in. thick with a 10 ft by 12 ft joint spacing without dowels. Performance: Excessive faulting and some longitudinal panel cracks resulted in replacement of this section in 2010. (c) and (d) MnRoad Cell 92. Bonded concrete overlay 6 in. thick with a 10 ft by 12 ft spacing with dowels. Performance: Longitudinal cracking in some panels but no faulting. Replaced in 2010. Photos: Tom Burnham, MnDOT, July 2008.

181 RIGID PAVEMENT BEST PRACTICES A recent summary report from MnRoad (2009) provides design recommendations for bonded concrete on HMA: “Under interstate traffic loads, the best performing and most economical test section at MnROAD has been the 6-inch–thick concrete over 7 inches of existing HMA, installed with 5 x 6-foot panels. This recommendation fol- lows the national trend toward 6-inch thick concrete overlays, placed with 6 x 6-foot panels on higher volume roadways.” Limited information on bonded CRCP overlays suggests they perform better than bonded concrete overlays over HMA for equal thicknesses, given performance data from Texas (Kim et al., 2007). Sections 4 in. thick located on I-610 containing wire mesh and materials with low CTE performed adequately for 20 years. The LTPP results for bonded concrete overlays over PCC provide mixed results. The preceding findings are supported by Harrington (2008), who states the following: • Use bonded overlays to “add structural capacity and/or eliminate surface distress when the existing pavement is in good structure condition. Bonding is essential, so thorough surface preparation is necessary before resurfacing.” • Use unbonded overlays “to rehabilitate pavements with some structural deteriora- tion. They are basically new pavements constructed on an existing, stable platform (the existing pavement).” Additional State Design and Construction Practices A best practices document by Tayabji and Lim (2007) overviewed a selection of de- sign, materials, and construction features for new concrete pavements for four state DOTs (Illinois, Minnesota, Texas, and Washington). These practices were updated based on recent information and are summarized in Tables 3.5 and 3.6. Minnesota and Washington were grouped together in Table 3.5 because their practices are for JPCP. Illinois and Texas are summarized in Table 3.6 to reflect their CRCP practices. Although these practices were developed with new pavement construction in mind, they are also applicable to long-life concrete overlay systems. A recurring theme emerges when examining these practices: (1) thick unbonded PCC slabs >11 in. are used, (2) design lives are all >30 years ranging up to 60 years, and (3) PCC mix and materials requirements are important. Thus, as expected, long- life PCC renewal options are not just about slab thickness, but also about materials and construction.

182 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 3.5. EXAMPLES OF LONG-LIFE JPCP STANDARDS FOR THE MINNESOTA AND WASHINGTON DOTS Item Minnesota DOT WSDOT Design life • 60 years • 50 years Typical structure • Slab thicknesses = 11.5–13.5 in. • 3–8-in. dense-graded granular base • Subbase 12–48 in. select granular (frost- resistant) • Slab thickness = 12–13 in. (typical) • 4-in. HMA base • 4-in. crushed stone subbase Joint design • Spacing = 15 ft with dowels • All transverse joints are doweled • Spacing = 15 ft with dowels • Joints saw-cut with single pass • Hot-poured sealant Dowel bars • Diameter = 1.5 in. (typical) • Length = 15 in. (typical) • Spacing = 12 in. • Bars must be corrosion resistant • Diameter = 1.5 in. • Length = 18 in. • Spacing = 12 in. • Bars must be corrosion resistant; epoxy coatings not acceptable Outside lane and shoulder • 14-ft lane with tied PCC or HMA • 12-ft lane with tied and dowel PCC Surface texture • Astroturf or broom drag • Longitudinal direction • Requires 1 mm average depth in sand patch test (ASTM E965) • Longitudinal texturing Alkali-silica reactivity (ASR) • Fine aggregate must meet ASTM C1260 (ASR Mortar-Bar Method) • Expansion ≤0.15% OK. If ≥0.30%, reject. • Mitigation required by use of GGBFS or fly ash when expansion is between 0.15 and 0.30% • Allow various combinations of Class F fly ash and ground granulated blast furnace slag (GGBFS) Aggregate gradation • Use a combined gradation • Use a combined gradation Concrete permeability • Use GGBFS or fly ash to lower permeability of concrete • Apply ASTM C1202 for rapid chloride ion permeability test Air content • 7.0% ± 1.5% • 5.5% Water/ cementitious ratio • ≤0.40 • ≤0.44 • Minimum cementitious content = 564 lb/yd3 of PCC mix Curing • No construction or other traffic for 7 days or flexural strength ≥350 psi • Traffic opening compressive strength ≥2,500 psi by cylinder tests or maturity method Construction quality • Monitor vibration during paving Source: Tayabji and Lim, 2007; Minnesota DOT, 2005a, 2005b; WSDOT, 2010.

183 RIGID PAVEMENT BEST PRACTICES CONCEPTS FOR DEVELOPING LONG-LIFE RENEWAL STRATEGIES Commonly accepted criteria for defining long-life concrete pavement performance ( Tayabji and Lim, 2007) were described previously. For the purposes of this document, those criteria are generally applicable, although the performance life requirement has been extended to 50 years. Long performance life, in combination with good ride quality and minimal dis- tress, cannot be achieved with increased pavement thickness or improved structural design alone. It requires the selection of durable component materials, proper mixture proportioning, comprehensive structural design, and best practices for construction to ensure acceptable long-term performance. Furthermore, it must be recognized that changes in one design or construction parameter (thickness or curing practices, for example) may have implications for the selection of other design parameters (joint spacing, for example). In other words, the pavement structure, materials, and construc- tion practices must be recognized as a system where the failure of any one component TABLE 3.6. EXAMPLES OF LONG-LIFE CRCP STANDARDS FOR THE ILLINOIS AND TEXAS DOTS Item Illinois DOT Texas DOT Design life • 30–40 years • 30 years Typical structure • Up to 14-in. CRCP slab • 4–6-in. HMA base • 12-in. aggregate subbase • Up to 13-in. CRCP slab with one layer of reinforcing steel • 14–15-in. CRCP slab with two layers of reinforcing steel • Uses stabilized base either 6-in. CTB with 1-in. HMA bond breaker on top or 4-in. HMA • Recommends tied PCC shoulders Tiebars • Use at centerline and lane-to-shoulder joints • Use 1 in. × 30 in. bars spaced at 24 in. CRCP reinforcement • Reinforcement ratio = 0.8% • Steel depth 4.5 in. for 14-in. slabs • All reinforcement in CRCP epoxy coated • Increased amount of longitudinal steel • Design details for staggering splices Aggregate requirements • Illinois DOT applies tests to assess aggregate freeze-thaw and ASR susceptibilities PCC mix • Limits the coefficient of thermal expansion of concrete to ≤6 microstrains per °F Construction requirements • Limits on concrete mix temperature = 50°F–90°F • Slipform pavers must be equipped with internal vibration and vibration monitoring • Curing compound must be applied within 10 min of concrete finishing and tining • Curing ≥7 days before opening to traffic • Revised construction joint details Source: Tayabji and Lim, 2007; TxDOT, 2009a, 2009b, 2011.

184 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE (whether structural, functional, or related to durability) results in a system that will not achieve the goal of long life. One general concept or approach for developing a long-life pavement design or renewal strategy is to identify potential failure mechanisms and address each of them in the design, construction, and/or materials specifications. There are many potential failure mechanisms that may limit the performance life of a given pavement structure, and each of these mechanisms can be addressed in the materials, design, and construc- tion specifications and procedures. Key considerations often include the following: • Foundation support (uniformity, volumetric stability, including stabilizing treatments); • Drainage design (moisture collection and removal and design for minimal maintenance); • Concrete mixture proportioning and components (e.g., selected to minimize shrinkage and potential for chemical attack, low CTE, provide adequate strength); • Dowels and reinforcing (corrosion resistance, sized and located for good load transfer); • Accuracy of design inputs; • Construction parameters (including paving operations, surface texture, initial smoothness); and • Quality assurance/quality control (QA/QC; e.g., certification, prequalification, inspection). All the potential failure mechanisms (including those associated with structural or functional deterioration) must be addressed to ensure the pavement system achieves the desired level of performance over 50 or more years. Addressing only one or two distresses or design parameters (e.g., only pavement slab thickness and joint spacing to reduce uncontrolled cracking) while ignoring others (such as durability of materials and concrete curing practices) may postpone the development of some distresses for 50 or more years without preventing the pavement from failing due to other distresses in less than 50 years. The overall pavement performance life will be only as long as the “weakest link” (or shortest life) in the chain of factors that controls the system. The need for a “systems approach” to long-life pavement renewal or design is illustrated in Figure 3.9. The chart presents an illustration of the expected performance life of an example standard pavement (with a 35-year nominal design life) due to the impacts of various design, materials, and construction parameters. It can be seen that, for this example, all of the components being considered result in a life of about 35 years; if we consider the pavement to be “failed” when any of the component perfor- mances “fails,” then the expected life of this pavement is equal to the shortest compo- nent performance life (about 28 years in this case, limited by the dowel bar corrosion). The chart in Figure 3.10 illustrates an effort to increase the pavement performance life to 50 years by improving several design and construction parameters (e.g., slab thickness, improved drainage and foundation support). Although the development of

185 RIGID PAVEMENT BEST PRACTICES Figure 3.9. Pavement designed and built for 35-year service life. 29 2014.05.10 03 R23 Guide Chapter 3-final for composition.docx “fails,” then the expected life of this pavement is equal to the shortest component performance life (about 28 years in this case, limited by the dowel bar corrosion). [Insert Figure 3.9] e 3.9. Pave signed and built for 35-year s rvice life. The chart in Figure 3.10 illustrates an effort to increase the pavement performance life to 50 years by improving several design and construction parameters (e.g., slab thickness, improved drainage and foundation support). Although the development of distresses due to these parameters is not expected to produce “failures” for at least 50 years, the overall pavement life remains controlled by the durability of the dowel bars. The goal of a 50-year performance life was not achieved. The chart in Figure 3.11 shows that the consideration of all of the potential improvement areas is necessary to ensure a performance life of at least 50 years. Figure 3.10. Improved design and construction specifications.

186 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE distresses due to these parameters is not expected to produce “failures” for at least 50 years, the overall pavement life remains controlled by the durability of the dowel bars. The goal of a 50-year performance life was not achieved. The chart in Figure 3.11 shows that the consideration of all of the potential improvement areas is necessary to ensure a performance life of at least 50 years. MATERIAL CONSIDERATIONS Although standard concrete pavement mixtures are suitable for the construction of unbonded concrete overlays, concrete is a complex material and involves judicious selection and optimization of various materials to produce a durable concrete (Van Dam et al., 2002). The concrete materials requirements reviewed largely focused on cementitious materials and aggregates. Figure 3.11. Illustration that all areas of improvement need to be considered for long life.

187 RIGID PAVEMENT BEST PRACTICES Cementitious Materials Cementitious materials include hydraulic cements, such as portland cement, and poz- zolanic materials, such as fly ash. Fly ash is also referred to as supplementary cementi- tious material (SCM). The current practice for paving concrete is to incorporate port- land cement and an SCM. Although not a common practice, some agencies allow use of ternary concrete mixtures that incorporate portland cement and two SCMs. Supplementary Cementitious Materials For highway paving applications, the choice of SCM is typically limited to fly ash and ground granulated blast furnace slag (GGBFS). The replacement dosage for SCMs (fly ash and GGBFS) should be compatible with the needs for strength and durability, with upper limits generally defined by state DOT standard specifications. For paving applications, the desired SCM content should be established considering durability concerns (ASR), if applicable, along with economic and sustainability considerations. Fly ash and slag are covered under the Environmental Protection Agency’s Com- prehensive Procurement Guidelines (CPG) (Environmental Protection Agency, 2011). The CPGs are federal laws that require federally funded construction projects to include certain recycled materials in construction specifications. Concrete specifica- tions, therefore, must include provisions that allow use of fly ash and slag. The CPGs state that no preference should be given to one of these materials over another; rather, they should all be included in the specification. The enabling federal legislation is from the Resource Conservation and Recovery Act (RCRA). Fly Ash Fly ash must meet the requirements of ASTM C618; however, care should be taken in applying ASTM C618 because it is rather broad. Class F fly ash is the preferred choice for controlling ASR, and it also improves sulfate resistance. Selection of fly ash type and dosage for ASR mitigation should be based on local best practices. A photo of Class F fly ash is shown in Figure 3.12. Typical dosages for Class F fly ash are generally between 15% and 25% by mass of cementitious materials. Sources must be evaluated for typical usage rates. As the amount of fly ash increases, some air-entraining and water-reducing admixtures are not as effective and require higher dosage rates due to interactions with the carbon in the fly ash. While ASTM C618 permits up to 6% loss on ignition (LOI), the state Figure 3.12. Class F fly ash. Photo: FHWA.

188 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE DOTs should establish their own LOI limits. Changes in LOI can result in changes to the amount of air-entraining admixture required in the mixture. If fly ash will be used to control expansion due to ASR, the lower the CaO content the more effective it will be. Ideally, the CaO content should not exceed 8%. Slag Cements and Ground Granulated Blast Furnace Slag (GGBFS) In the recent past, cement typically used in concrete pavements was traditional port- land cement Type I or II (or occasionally Type III for decreased cure times). Today, a wider range of cements is available, including slag cements and cements that are com- binations of portland and slag cement. Blast furnace slag is a by-product of manufacturing molten iron in a blast furnace. This granular material (Figure 3.13) results when the molten slag is quenched with water. The rapid cooling forms glassy silicates and aluminosilicates of calcium. Once ground to a suitable particle size, the end result is GGBFS. This is commonly referred to as “slag cement.” GGBFS must meet the requirements of ASTM C989. The following three grades are based on their activity index: 1. Grade 80 is the least reactive and is typically not used for highway or airport projects. 2. Grade 100 is moderately reactive. 3. Grade 120 is the most reactive, with increased activity achieved through finer grinding. Grade 120 can be difficult to obtain in some regions of the United States. It is common that blends of slag and portland cements are made (typically desig- nated Type IS(X), where X is the percentage of GGBFS). Typical dosages of slag should be between 25% and 50% of cementitious materials. Concrete strength at early ages (up to 28 days) may be lower using slag–cement combinations, particularly at low Figure 3.13. Preprocessed blast furnace slag. Photos: Joe Mahoney.

189 RIGID PAVEMENT BEST PRACTICES temperatures or at high slag percentages. The desired slag content must be established by considering the importance of early strengths for the panel-fabrication process. However, if the slag will be used to control expansions caused by ASR, the minimum slag content used is that needed to control ASR. Aggregates Aggregates are a key component of concrete and can affect the properties of both fresh and hardened concrete. This is, in part, due to 70% to 80% of the PCC volume being composed of aggregates. Aggregate selection should maximize the volume of aggregate in the concrete mixture to minimize the volume of cementitious paste (without com- promising the durability and strength of the concrete mixture). Aggregate requirements for pavement concrete are typically established in accordance with the requirements of ASTM C33. Some of the key aggregate requirements are discussed below. Tables 3.7 and 3.8 summarize the relationship between aggregate properties and possible pave- ment distresses and standard test methods (Folliard and Smith, 2003) and illustrate the critical roles of competent aggregates. Figure 3.14 shows typical aggre gate processing before batching concrete for paving. Figure 3.14. Aggregate processing, which includes stockpiles, conveyors, and screening. Photos: Joe Mahoney.

190 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 3.7. CONCRETE PAVEMENT PERFORMANCE PARAMETERS AFFECTED BY AGGREGATE PROPERTIES Performance Parameter Manifestation Mechanism(s) PCC Properties Aggregate Properties Alkali-aggregate reactivity Shallow map cracking and joint/crack spalling, accompanied by staining Chemical reaction between alkalis in cement paste and either susceptible siliceous or carbonate aggregates • Mineralogy • Size • Porosity Blowups Upward lifting of PCC slabs at joints or cracks, often accompanied by shattered PCC Excessive expansive pressures caused by incompressibles in joints, alkali- aggregate reactivity (AAR), or extremely high temperature or moisture conditions • Coefficient of thermal expansion • Coefficient of thermal expansion • Mineralogy D-cracking Crescent-shaped hairline cracking generally occurring at joints and cracks in an hourglass shape Water in aggregate pores freezes and expands, cracking the aggregate and/or surrounding mortar • Air void quality • Mineralogy • Pore size distribution • Size Longitudinal cracking Cracking occurring parallel to the centerline of the pavement Late or inadequate joint sawing, presence of alkali-silica reactivity (ASR), expansive pressures, reflection cracking from underlying layer, traffic loading, loss of support • Coefficient of thermal expansion • Coarse aggregate– mortar bond • Shrinkage • Coefficient of thermal expansion • Gradation • Size • Mineralogy • Shape, angularity, and texture • Hardness • Abrasion resistance • Strength Roughness Any surface deviations that detract from the rideability of the pavement Development of pavement distresses, foundation instabilities, or “built in” during construction • Any that affect distresses • Elastic modulus Workability • Any that affect distresses • Gradation • Elastic modulus Spalling Cracking, chipping, breaking, or fraying of PCC within a few feet of joints or cracks Incompressibles in joints, D-cracking or AAR, curling/warping, localized weak areas in PCC, embedded steel, poor freeze-thaw durability • Coefficient of thermal expansion • Coarse aggregate– mortar bond • Workability • Durability • Strength • Air void quality • Shrinkage • Gradation • Mineralogy • Texture • Strength • Elastic modulus • Size continued

191 RIGID PAVEMENT BEST PRACTICES Performance Parameter Manifestation Mechanism(s) PCC Properties Aggregate Properties Surface friction Force developed at tire–pavement interface that resists sliding when braking forces applied Final pavement finish and texture of aggregate particles (mainly fine aggregates) • Hardness • Shape, angularity, and texture • Mineralogy • Abrasion resistance Transverse cracking Cracking occurring perpendicular to the centerline of the pavement PCC shrinkage, thermal shrinkage, traffic loading, curling/ warping, late or inadequate sawing, reflection cracking from underlying layer, loss of support • Shrinkage • Coarse aggregate– mortar bond • Coefficient of thermal expansion • Strength • Coefficient of thermal expansion • Gradation • Size • Shape, angularity, and texture • Mineralogy • Hardness • Abrasion resistance • Strength Corner breaks (jointed PCC) Diagonal cracks occurring near the juncture of the transverse joint and the longitudinal joint or free edge Loss of support beneath the slab corner, upward slab curling • Strength • Coarse aggregate– mortar bond • Coefficient of thermal expansion • Elastic modulus • Coefficient of thermal expansion • Gradation • Size • Mineralogy • Shape, angularity, and texture • Hardness • Abrasion resistance • Strength Transverse joint faulting (jointed PCC) Difference in elevation across transverse joints Pumping of fines beneath approach side of joint, settlements or other foundation instabilities • Elastic modulus • Size • Gradation • Shape, angularity, and texture • Abrasion resistance • Elastic modulus • Coefficient of thermal expansion Punchouts (CRCP) Localized areas of distress characterized by two closely spaced transverse cracks intersected by a longitudinal crack Loss of support beneath slab edges and high deflections • Elastic modulus • Strength • Shrinkage • Coefficient of thermal expansion • Elastic modulus • Strength • Coefficient of thermal expansion • Size • Shape, angularity, and texture • Abrasion resistance Source: After Folliard and Smith, 2003. TABLE 3.7. CONCRETE PAVEMENT PERFORMANCE PARAMETERS AFFECTED BY AGGREGATE PROPERTIES (continued)

192 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 3.8. STANDARD AGGREGATE, AGGREGATE-RELATED, AND PCC TEST METHODS Property Test Method Basic aggregate property Grading AASHTO T27 Specific gravity AASHTO T84 Absorption AASHTO T84 Unit weight AASHTO T19 Petrographic analysis ASTM C295 Durability Soundness AASHTO T104 F-T resistance AASHTO T161 Internal pore structure AASHTO T85 Degradation resistance AASHTO T96, ASTM C535 Chemical reactivity ASR ASTM C227, C295, C289 Alkali-carbonate reactivity (ACR) ASTM C295 Dimensional change Drying shrinkage ASTM C157 Deleterious substances AASHTO T21 Frictional resistance AASHTO T242 Particle shape and texture ASTM D4791 Source: Folliard and Smith, 2003. Maximum Aggregate Size The concern with aggregate size involves selecting an aggregate that will maximize aggre gate volume and minimize cementitious material volume. In general, the larger the maximum size of the coarse aggregate, the less cementitious material is required, potentially leading to lower costs. Use of smaller maximum-size aggregate (e.g., 0.75-in. maximum size) is required for D-cracking regions. However, the use of 0.75-in. maxi- mum aggregate size alone does not prevent D-cracking, and many state agencies have criteria for D-cracking other than maximum aggregate size. Aggregate Gradation In the past, paving concrete was produced using coarse and fine aggregates. Today, agencies are moving toward the use of a combined gradation that may require use of more than two aggregate sizes. A combined gradation is based on an 8-to-18 speci- fication. The percentage retained on all specified standard sieves should be between 8% and 18%, except for the coarsest sieve and sieves finer than the No. 30 sieve. The coarseness factor differentiates between gap-graded and well-graded aggregate grada- tions, whereas the workability factor determines the mix coarseness. Concrete made with combined aggregate gradation has improved workability for slipform paving appli cations, requires use of less cementitious materials, exhibits less drying shrinkage, and may be more economical (Richardson, 2005).

193 RIGID PAVEMENT BEST PRACTICES Deleterious Substances Deleterious substances are contaminants that are detrimental to the aggregate’s use in concrete. ASTM C33 lists the following as deleterious substances: • Clay lumps and friable particles, • Chert (with saturated surface dry specific gravity <2.40), • Material finer than a No. 200 sieve, and • Coal and lignite. Inclusion of larger-than-allowable amounts of the deleterious substances can seri- ously impact both the strength and durability of concrete. Soundness The soundness test measures the aggregate’s resistance to weathering, particularly frost resistance. The ASTM C88 test for soundness has a poor precision record. Aggregates that fail this test may be reevaluated using ASTM C666 or judged on the basis of local service history. Flat and Elongated Particles Flat and elongated particles affect the workability of fresh concrete and may negatively affect the strength of hardened concrete. The amount of such particles needs to be limited. The breakdown of aggregates, especially the breakdown of fine aggregates, during handling and later when mixed in the concrete may lead to the production of excess microfines. This aggregate breakdown tends to negatively affect concrete workability, its ability to entrain air, and constructability (i.e., placing, compacting, and finishing). Increasing water content to offset the reduction in workability would increase the w/c ratio and lead to lower strength and an increased potential of plastic and drying shrinkage (Folliard and Smith, 2003). Los Angeles Abrasion Test The Los Angeles abrasion test provides a relative assessment of the hardness of the aggregate. Harder aggregates maintain skid resistance longer and provide an indicator of aggregate quality. Durability (D-Cracking) Durability cracking (D-cracking) is a concern for coarse aggregate particles that typi- cally are (1) sedimentary in origin, (2) have a high porosity, (3) have small pore size (about ∼0.1 µm), and (4) become critically (>91%) saturated and subjected to freezing and thawing. Cracking of the concrete is caused by the dilation or expansion of sus- ceptible aggregate particles and will develop wherever the conditions of critical satura- tion and freezing conditions exist. Because moisture is usually more readily available near pavement joints and cracks, patterns of surface cracking often surround and fol- low the joints and cracks, as shown in Figure 3.15. Also, because there is usually more moisture present at the bottom of the slab than at the surface, the extent of cracking deterioration is often much greater than what is visible at the surface.

194 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Van Dam et al. (2002) hypothesized that D-cracking is caused by aggregates with a certain range of pore sizes, and the damage may be exacerbated in the presence of deicing salts for some carbonate aggregates. Coarse aggregates are the primary con- cern, and for each specific aggregate type, there generally exists a critical aggregate size below which D-cracking is not a problem. Coarse aggregate particles exhibiting relatively high absorption and having pore sizes ranging from 0.1 to 5 µm generally experience the most freezing and thawing problems because of higher potential for sat- uration. Aggregates of sedimentary origin, such as limestones, dolomites, and cherts, are most susceptible to D-cracking (Van Dam et al., 2002). Alkali-Aggregate Reactivity (AAR) Two types of AAR reaction are recognized, and each is a function of the reactive mineral; silicon dioxide or silica (SiO2) minerals are associated with ASR and calcium magnesium carbonate [CaMg(CO3)2 or dolomite] minerals with alkali-carbonate reac- tivity (ACR) (Thomas, Fournier, and Folliard, 2008). Both types of reaction can result in expansion and cracking of concrete elements, leading to a reduction in the service life of concrete structures. A process for identifying whether there is (or could be) a problem with AAR is illustrated in Figure 3.16. ASR is of more concern because the aggregates associated with it are common in pavement construction. ASR is a deleterious chemical reaction between reactive silica constituents in aggregates and alkali hydroxides in the hardened cement paste. This constituent of concrete has a pore structure, and the associated pore water is an alkaline solution. This alkaline condition, plus reactive silica provided by the aggre- gate, produces a gel. The gel, unfortunately, has an affinity for water, which in turn grows and produces expansive stresses. These stresses generate polygonal cracking within the aggregate, within the mortar, or both that over time can compromise the structural integrity of concrete. Concrete undergoing ASR often exhibits telltale signs of surface map cracking as illustrated by Figures 3.17 and 3.18. It is widely accepted Figure 3.15. D-cracking. Photos: FHWA, NHI.

195 RIGID PAVEMENT BEST PRACTICES that high-pH (>13.2) pore water in combination with an optimum amount of reactive siliceous aggregate are key ingredients to initiate ASR expansion; it is also believed that a relative humidity (RH) ≥85% is essential for ASR to occur. Although the problem is widely known, and successful mitigation methods are available, ASR continues to be a concern for concrete pavement. Aggregates suscep- tible to ASR are either those composed of poorly crystalline or metastable silica mate- rials, which usually react relatively quickly and result in cracking within 5 to 10 years, or those involving certain varieties of quartz, which are slower to react in field applica- tions. ASR research is ongoing and the provisions associated with ASR-related test- ing are based on best current practices. Guidelines related to ASR will continue to be updated or replaced as more research becomes available. AASHTO has issued a Provisional Practice—AASHTO Designation PP 65-10—to address ASR. The full title of PP 65-10 is “Provisional Practice for Determining the Reactivity of Concrete Aggregates and Selecting Measures for Preventing Deleterious Figure 3.16. Evaluation stages for alkali-aggregate reaction determination. Source: Thomas et al., 2008.

196 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Figure 3.17. Illustration of ASR on a traffic barrier. Photo: FHWA. Figure 3.18. Illustration of ASR in concrete pavements. Source: D. Huft, South Dakota DOT. Expansion in New Concrete Construction.” Additionally, reports from PCA (Farney and Kosmatka, 1997) and FHWA (Thomas et al., 2008: Fournier et al., 2010) provide solid explanations of why ASR occurs, how it can be assessed, and mitigation mea- sures that can be taken. Coefficient of Thermal Expansion The coefficient of thermal expansion (CTE) plays an important role in PCC joint design (including joint width and slab length) and in accurately computing pavement stresses (especially curling stresses) and joint load-transfer efficiency (LTE) over the design life; thus, the lower the CTE the better for concrete pavements.

197 RIGID PAVEMENT BEST PRACTICES The CTE of concrete is highly dependent on the CTEs of the concrete components and their relative proportions (as well as the degree of saturation of the concrete). Cement-paste CTE increases with water-to-cement ratio, and cement pastes generally have higher CTEs than concrete aggregates (as shown in Table 3.9). Therefore, the concrete aggregate, which typically comprises 70% or more of the volume of concrete, tends to control the CTE of the hardened concrete: more aggregate and lower CTE aggregate results in concrete with lower CTE values. It should be noted that critical internal stresses may develop in the PCC if the thermal expansion characteristics of the matrix and the aggregates are substantially different, and large temperature changes take place. TABLE 3.9. TYPICAL CTE RANGES FOR COMMON PCC COMPONENTS Material Type Typical CTE (× 10−6/oF) Aggregate Limestone 3.4–5.1 Granites and gneisses 3.8–5.3 Basalt 4.4–5.3 Dolomites 5.1–6.4 Sandstones 5.6–6.5 Quartz sands and gravels 6.0–8.7 Quartzite, cherts 6.6–7.1 Cement paste with w/c ratio 0.4–0.6 10.0–11.0 Concrete cores from LTPP sections 4.0 (lowest), 5.5 (mean), 7.2 (highest) Source: ARA, 2004. Chemical Admixtures A number of chemical admixtures can be added to concrete during proportioning or mixing to enhance the properties of fresh and/or hardened concrete. Admixtures commonly used in mixtures include air entrainers and water reducers. The stan- dard specification for chemical admixtures in concrete used in the United States is AASHTO M194 (ASTM C494). The use of chemical admixtures for concrete is a well- established practice and requires no additional provisions for application. High-range water reducers are typically not used with paving concrete. Other Materials The characteristics of other materials used in the construction of unbonded concrete overlays are as follows: • Dowel bars should conform to the appropriate ASTM and AASHTO standards. The standard practice in the United States is to specify use of epoxy-coated dowel bars. However, the effectiveness of the current standard epoxy coating mate rials and processes beyond 15 to 25 years in service is considered suspect. Figure 3.19 shows epoxy-coated dowels with less than 15 years of service in Washington State.

198 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE It is noted that these photos are from retrofit dowel projects, which present chal- lenges in consolidating the patching mix—a situation unlikely to occur in PCC overlays; however, voids in the vicinity of dowels are a concern. Corrosion has been noted for epoxy-coated dowels by WSDOT on fully reconstructed JPCP con- struction following about 15 years of service. Several recent projects (in Minnesota, Illinois, Iowa, Ohio, and Washington State) have been constructed using stainless steel–clad dowel bars (Figure 3.20) and zinc-clad dowel bars with satisfactory performance (Federal Highway Administration, 2006). WSDOT requires corro- sion-resistant dowel bars for concrete pavements that have a design life of greater than 15 years. The long-life dowel options used by WSDOT include (1) stainless steel–clad bars, (2) stainless steel tube bars whereby the tube is press-fitted onto a plain steel inner bar, (3) stainless steel solid bars, (4) corrosion-resistant steel bars that conform to ASTM A1035, and (5) zinc-clad bars (Washington State Depart- ment of Transportation, 2010). The Minnesota and Wisconsin DOTs have similar specifications for long-life dowel bars, with Minnesota allowing the use of hollow stainless steel tubes as an additional option, and neither state allowing the A1035 dowels (Minnesota Department of Transportation, 2005b; Wisconsin Department of Transportation, 2009). Additional guidance on dowel bar design can be found in a recent publication by the Concrete Pavement Technology Center (2011). • Tiebars should conform to the appropriate ASTM and AASHTO standards. • All joint cuts and sealant materials used should conform to the appropriate ASTM and AASHTO standards, or a governing state specification. Figure 3.19. Corroded epoxy-coated dowel bars in a retrofitted dowel bar project (original bars 1.5 in. by 18 in.). Photos: WSDOT.

199 RIGID PAVEMENT BEST PRACTICES UNBONDED CONCRETE OVERLAYS OF CONCRETE PAVEMENTS Criteria for Long-life Potential This renewal strategy is applicable when the existing pavement exhibits extensive structural deterioration and possible material-related distresses such as D-cracking or reactive aggregate (Smith, Yu, and Peshkin, 2002: Harrington, 2008). The success of the strategy depends on the stability (structural integrity) and the uniformity of the under lying structure. Since the concrete overlay is “separated” from the underlying pavement, the preoverlay repairs are usually held to a minimum. Figure 3.21 is a sketch of an unbonded overlay over concrete. Figure 3.22 illustrates an in-service unbonded undoweled concrete overlay. The photo shows a 35-year-old JPCP overlay over an existing JPCP located on I-90 in Washington State. The following sections summarize some of the design and construction issues to consider for long-life unbonded concrete overlays. Figure 3.20. Stainless steel dowel bar. Photo: Joe Mahoney.

200 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Figure 3.21. Unbonded concrete overlay of concrete pavement. Illustration: Joe Mahoney. Figure 3.22. Unbonded 9-in. JPCP concrete overlay placed over concrete (I-90 in Washington State; overlay is 35 years old). Photo: WSDOT. Joints Unbonded Concrete Overlay Interlayer Existing Concrete Pavement

201 RIGID PAVEMENT BEST PRACTICES General Design Considerations Smith et al. (2002) and Harrington (2008) have suggested that, when designing un- bonded concrete overlays, the following factors need to be considered: • The type and condition of the existing pavement. In general, unbonded concrete overlays are feasible when the existing pavement is in poor condition, including material-related distress such as sulfate attack, D-cracking, and ASR. The struc- tural condition of the existing pavement can be established by (1) conducting visual distress surveys, (2) conducting deflection testing using a falling weight deflectometer (FWD) (the deflection magnitudes can be used to determine the load-transfer effi ciency across joints, determine possible support characteristics under the slab corners and edges, back-calculate the modulus of subgrade reac- tion and modulus of the existing portland cement concrete pavement, and deter- mine the variability of the foundation layers along the length of the project); and (3) extracting cores from the existing pavement. Laboratory testing of the cores is necessary if the existing pavement exhibits D-cracking or reactive aggregates. • Preoverlay repairs. One of the attractive features of this renewal strategy is that extensive preoverlay repairs are not warranted. It is recommended that only those distresses need to be addressed that can lead to a major loss in structural integrity and uniformity of support. The guidelines (Harrington, 2008) for conducting pre- overlay repairs are summarized in Table 3.10. TABLE 3.10. GUIDELINES FOR PREOVERLAY REPAIRS Existing Pavement Condition Possible Repairs Faulting ≤10 mm No repairs needed Faulting >10 mm Use a thicker interlayer Significant tenting, shattered slabs, pumping Full-depth repairs Severe joint spalling Clean the joints CRCP with punchouts Full-depth repairs Source: Harrington, 2008. • Separator-layer design. The separator layer is a critical factor for the performance of the unbonded concrete overlay. The separator layer acts as a lower-modulus buffer layer that assists in mitigating cracks from reflecting up from the existing pavement to the new overlay. The separator layer does not contribute significantly to the structural enhancement. Structural Design and Joint Design Considerations The design thickness of unbonded PCC overlays is typically ≤≤9 in. for Interstate appli- cations. Figure 3.23 illustrates the probability of poor performance of unbonded con- crete overlays in these applications as a function of slab thickness. It is evident that, for long-life pavements in high-traffic-volume applications, the overlay thickness should

202 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE be 9 in. or greater. It is clear that slab thickness is one of the critical design features for ensur ing long service life; however, the slab thickness required for long pavement life may vary somewhat with other design details (e.g., joint design and layout), and long life cannot be achieved at any slab thickness unless sufficiently durable materials are used. Thickness design can be performed using either the AASHTO 1993 or the Mechanistic-Empirical Pavement Design Guide (MEPDG) design methods. The key factors associated with these two methods are described below: In the AASHTO design method (1993/1998), the overlay design is based on the concept of structural deficiency, in which the structural capacity of the unbonded con- crete overlay is computed as a difference between the structural capacity of the new pavement designed to carry the projected traffic and the effective structural capacity of the existing pavement. The effective structural capacity of the existing pavement can be established using (1) the condition survey method or (2) the remaining life method. The thickness of the new pavement required to carry the projected traffic can be determined by using the AASHTO design procedure for new PCC pavements. This method of design does not take into account the interaction (friction and bonding) between the separator layer and the overlay and separator layer and the existing pave- ment. The 1993/1998 AASHTO overlay design method does not directly account for the effects of thermal (curling) and moisture (warping) gradients. The results tend to Figure 3.23. Slab thickness versus probability of poor performance for unbonded JPCP overlays. Source: Smith et al., 2002.

203 RIGID PAVEMENT BEST PRACTICES be conservative for high-ESAL conditions and often calculate greater concrete overlay design thicknesses than mechanistic-based procedures. The MEPDG (or Darwin-ME) design method is based on the damage concept and uses an extensive array of inputs to estimate pavement distress for a specific set of inputs. The predicted distress types for JPCP are slab cracking, faulting, and IRI. For CRCP, the predicted distress types are punchouts and IRI. The production version of the MEPDG (Darwin-ME) from AASHTO was released during 2011. Joint design is one of the factors affecting jointed pavement performance. It also affects the thickness design for overlays. The joint design process includes joint spac- ing, joint width, and load-transfer design (dowel bars and tiebars). Size, layout, and coating of the dowel bars depend on the project location and traffic levels. Load transfer in unbonded concrete resurfacing is typically very good— comparable to that of new JPCP on HMA base, and better than that of JPCP on untreated base. Doweled joints should be used for unbonded resurfacing on pavements that will expe- rience significant truck traffic (i.e., typically for concrete overlay thicknesses of 9 in. or more). Several studies have shown that adequately sized dowels must be provided to obtain good faulting performance (Snyder et al., 1989; Smith et al., 1997). Dowel diameter is often selected based on slab thickness, but traffic may be a more important factor for consideration. For long-life pavements, 1.5-in.-diameter bars are usually recommended. Additionally, corrosion-resistant dowels (e.g., stainless steel–surfaced, nonstainless corrosion-resistant steel (ASTM A1035), and zinc-clad steel alternatives are required by those state DOTs considering long-life designs. Details concerning the design of dowel load-transfer systems can be found in a recent publication prepared by the National Concrete Consortium (Concrete Pavement Technology Center, 2011). Examples of three state DOT specifications and special provisions for the use of corro- sion-resistant dowels were cited earlier. Figure 3.24. Joint mismatching details. Source: Smith et al., 2002.

204 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE It is recommended that shorter joint spacings be used to reduce the risk of early cracking due to curling stresses. A maximum joint spacing of 15 ft is typically used for thick (>9 in.) long-lived concrete pavements. Figure 3.24 illustrates a typical joint mismatching detail, which should be considered for jointed concrete overlays. Prior recommendations suggest that the transverse joints should be sawed to a depth of T/4 (minimum) to T/3 (maximum) (Smith et al., 2002; Harrington, 2008). Drainage Design Drainage system quality significantly affects pavement performance. Overlay drain- age design depends on the performance and capacity of the existing drainage system. Consequently, evaluation of the existing pavement is the first step in overlay drainage design. Depending on the outcome of this evaluation, no upgrade may be necessary. However, in the presence of distresses caused by moisture, appropriate design mea- sures must be employed to address these issues. Distresses such as faulting, pumping, and corner breaks could be indicators of a poor drainage system. Standing water might be an indication of insufficient cross slope. Proper design, along with good construc- tion and maintenance, will reduce these types of distresses. If asphalt interlayer drain- age is inadequate in an unbonded PCC overlay, pore pressure induced by heavy traffic may cause HMA layer stripping, so careful consideration and design for interlayer drainage should be followed (Smith et al., 2002; Harrington, 2008). Separator Layers The separator layer is a critical factor in determining the performance of an unbonded concrete overlay. The separator layer acts as a lower-modulus buffer layer that assists in preventing cracks from reflecting up from the existing pavement to and through the new overlay. The separator layer does not contribute significantly to the structural enhancement and, therefore, the use of excessively thick (e.g., >2 in.) separator layers should be avoided (Smith et al., 2002; Harrington, 2008). Interlayers should be between 1 and 2 in. thick (Smith et al., 2002; Harrington, 2008). Thin interlayers (e.g., 1 in.) have been used successfully when the existing pave- ment has little faulting or other surface distress. Thicker separator layers have been used when faulting and distress levels are high. The use of dense-graded and permeable HMA interlayers is common. Other materials used in unbonded overlay interlayers (either alone or in conjunction with HMA material) include polyethylene sheeting, liquid asphalts, geotextile fabrics, chip seals, slurry seals, and wax-based curing com- pounds. Not all of these materials and material combinations may be suitable for long-life pavements. In Germany, a nonwoven fabric material is placed between the stabilized subbase and concrete slab to prevent bonding between layers and to provide a medium for subsurface drainage. This technology has been adapted for use in the United States for unbonded concrete overlay interlayers and was showcased on a 2008 unbonded concrete overlay project in Missouri (Tayabji et al., 2009). Figure 3.25 illustrates the placement of the fabric on the existing pavement surface. It is noted that no long-term performance data are currently available for the application of this technology in con- crete overlays.

205 RIGID PAVEMENT BEST PRACTICES Table 3.11 summarizes the types of interlayers currently used in the construction of unbonded concrete overlays for concrete pavements. This information is based on extended meetings with pavement engineering and management professionals from the Illinois Tollway Authority and the Michigan, Minnesota, and Missouri DOTs. As reported by Smith et al. (2002), the most commonly used separator layer is HMA (69%). Although other types of separator layers are also used, bituminous materials make up 91% of all separator layer types. TABLE 3.11. EXAMPLE STATE OF PRACTICE REGARDING THE USE OF INTERLAYERS State DOT Interlayer Material Illinois Tollway Authority Used rich sand asphalt layer for one project. Michigan Experienced problems with thick sandy layers. Moved to using open-graded interlayer with a uniform thickness. The HMA separation layer is constructed in either a uniform 1-in. or 1- to 3-in. moderately wedged section. Geometric issues are corrected with the thickness of the PCC overlay. Minnesota Typically use an open-graded interlayer, but have also milled existing HMA to a 2-in. thickness and utilized it as an interlayer. Missouri Typically use a 1-in. HMA or geotextile interlayer. Figure 3.25. Placement of nonwoven fabric as an interlayer. Source: Tayabji et al., 2009.

206 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Performance Considerations The performance of unbonded concrete overlays of GPS-9 sections is presented in this section. The pavement performance criteria selected for the summary include trans- verse cracking, IRI (and PSI), and joint and crack faulting. The performance trends presented in this section are based on measurements documented in the latest year of monitoring available. Transverse Cracking Figure 3.26 shows typical transverse cracks for both airfield and highway pavements. Figure 3.27 shows the magnitude of the average number of transverse cracks per 500-ft-long section for the LTPP GPS-9 sections as a function of overlay thickness for jointed concrete pavements. As expected, the thicker overlays (>8 to 9 in.) exhibit fewer transverse cracks. It is noted that 11 of the 14 jointed concrete pavement over- lays exhibited little or no cracking in 18 years of service. These test sections do exhibit the promise of long-life performance. International Roughness Index (IRI) Figure 3.28 illustrates the progression of IRI and PSI for the various GPS-9 sections and the impact of overlay thickness on ride quality. Joint and Crack Faulting Figure 3.29 illustrates transverse contraction joint faulting (faulting above 0.25 in. is significant); however, the data from GPS-9 projects do not show the degree of severity that is illustrated in Figure 3.30. The overall magnitude of the faulting is below 0.25 in. and therefore does not appear to be an issue; however, slab thicknesses greater than 9.6 in. show significantly less faulting, perhaps due to the use of dowel bars in these thicker pavements. The thinner overlays in the GPS-9 experiment were not dow- eled, so the trends are probably more due to the use of dowels rather than pavement Figure 3.26. Transverse cracking on an airport apron and an Interstate highway. Photos: Joe Mahoney.

207 RIGID PAVEMENT BEST PRACTICES 0 2 4 6 8 10 12 14 5.1" - 6.5" 6.6" - 8" 8.1" - 9.5" 9.6" - 11" Overlay Thickness Av g. N o. o f T C (L as t S ur ve y) Figure 3.27. JPCP overlay thickness versus average number of transverse cracks. 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 5.1" - 6.5" 6.6" - 8" 8.1" - 9.5" 9.6" - 11" Overlay Thickness Av g. IR I ( m /k m ) 0.0 1.0 2.0 3.0 4.0 5.0 Av g. P SI IRI PSI Figure 3.28. Overlay thickness versus average IRI and average PSI (pavement age ranges from 6 to 20 years).

208 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE 0.00 0.50 1.00 1.50 2.00 2.50 3.00 5.1" - 6.5" 6.6" - 8" 8.1" - 9.5" 9.6" - 11" Overlay Thickness Av g. W he el pa th F au lti ng (m m ) Figure 3.29. Overlay thickness versus average wheelpath faulting. (a) (b) Figure 3.30. Contraction joint faulting of JPCP. (a) Average fault ∼0.25–0.5 in. (b) Average fault ∼0.5 in. Photos: WSDOT. thickness, but that may simply imply that the pavement needs to be thick enough to install dowels. The use of properly designed dowels in the transverse joints should es- sentially eliminate transverse joint faulting. Impact of Interlayer Design on Performance Figures 3.31 and 3.32 illustrate the impact of the interlayer type and thickness on transverse cracking of the overlay. In general, thicker interlayers tend to inhibit trans- verse cracking.

209 RIGID PAVEMENT BEST PRACTICES 0 5 10 15 20 25 30 Dense Graded Asphalt Concrete Open Graded Asphalt Concrete Chip Seal Other No Interlayer Interlayer Type Av g. N o. o f T C (L as t S ur ve y) 0 2 4 6 8 10 12 14 0" 0.1" - 1.9" 2" - 3.8" > 3.9" Interlayer Thickness Av g. N o. o f T C (L as t S ur ve y) Figure 3.31. JPCP interlayer type versus average number of transverse cracks. Figure 3.32. JPCP interlayer thickness versus average number of transverse cracks.

210 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Figure 3.33 shows that thicker interlayers contribute to the integrity of the joint by controlling the amount of joint faulting (all other parameters being equal). Construction Considerations Construction of the Separator Layer The placement of a separator layer is straightforward. The procedure depends on the interlayer material, but standard application procedures apply. The existing pavement surface needs to be swept clean of any loose materials. Either a mechanical sweeper or an air blower may be used (American Concrete Pavement Association, 1990; McGhee, 1994). With HMA separator layers, precautionary steps may be needed to prevent the development of excessively high surface temperatures prior to PCC placement. Surface watering should be used when the temperature of the asphalt separator layer is at or above 120oF to minimize the potential of early age shrinkage cracking (Harrington, 2008). There should be no standing water or moisture on the separator-layer surface at the time of overlay placement. An alternative to this is to construct the PCC overlay at night. Whitewashing of the bituminous surface using lime slurry may also be per- formed to cool the surface (American Concrete Pavement Association, 1990). How- ever, this practice may lead to more complete debonding between the overlay PCC and the separator layer. Some degree of friction between the overlay PCC and the separator layer is believed to be beneficial to the performance of unbonded overlays, even if the Figure 3.33. JPCP interlayer thickness versus average wheelpath faulting. 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 0" 0.1" - 1.9" 2" - 3.8" > 3.9" Interlayer Thickness Av g. W he el pa th F au lti ng (m m )

211 RIGID PAVEMENT BEST PRACTICES structural design is based on the assumption of no bond (ERES, 1999). The size of the project and geometric constraints will determine the type of paving (fixed form, slip form, or a combination) used (Smith et al., 2002). Concrete Temperature During Construction During construction, excessively high temperature and moisture gradients through the PCC must be avoided through the use of good curing practices (i.e., control of concrete temperature and moisture loss). Several studies have shown that excessive temperature and/or moisture gradients through the PCC slab at early ages (particularly during the first 72 hours after placement) can induce a significant amount of curling into PCC slabs, which can then result in higher slab stresses and premature slab cracking. This built-in construction curling is of particular concern for unbonded overlays because of the very stiff support conditions typically present. Early age (less than 72 hours) characterization of the pavement should be per- formed to study the impact of PCC mixture characteristics and climatic conditions at the time of construction on the predicted overlay behavior and performance. An excel- lent tool for completing concrete pavement early age assessments is the HIPERPAV III software (High Performance Concrete Paving) (HIPERPAV, 2010). A screen shot from HIPERPAV is shown in Figure 3.34, which illustrates the predicted tensile stress and strength in the concrete over the first 72 hours following placement. Figure 3.34. Screen shot from HIPERPAV III software illustrating tensile stress and strength over first 72 hours. Source: HIPERPAV, 2010.

212 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Surface Texture For quieter pavements, the surface texture should be negative (i.e., grooves pointing downward, not fins) and oriented longitudinally. If the texture is placed in the trans- verse direction, then it should be closely spaced and randomized. Texture depth is also important for both friction and noise generation. A minimum depth is required for friction, but excessive depth of texture (particularly for transversely oriented textures) is associated with significantly greater noise generation, both inside and outside of the vehicle (American Concrete Pavement Association, 2006). It is believed that the use of siliceous sands tends to improve texture durability and friction. For diamond grinding, polish-resistant, hard and durable coarse aggregates are recommended. Narrow single- cut joints are recommended to minimize noise. Avoid faulted joints, protruding joint sealants, and spalled joints for quieter pavements (Rasmussen et al., 2008). Dowel Placement The use of dowel bars is critical for long-lasting JPCP. Numerous studies, including the AASHO Road Test, showed the need for doweled transverse contraction joints to survive heavy traffic conditions. A number of state DOTs during the initial construc- tion of the Interstate system used undoweled JPCP and have now changed to doweled JPCP—largely due to faulting of the contraction joints. During construction, dowel misalignment can occur, particularly so with dowel bar inserters—although it can hap- pen with dowel baskets as well. It is critical to avoid such misalignments, and technol- ogy developed over the past 10 years can help do so. There are five possibilities for misalignment, as illustrated in Figure 3.35. These misalignments can cause various types of performance issues ranging from slab spall- ing to cracking, as shown in Table 3.12. Notably, the long-term load transfer at the contraction joints can also be affected. As shown in the table, horizontal skew and vertical tilts are likely the most critical misalignments. Figure 3.35. Types of dowel bar misalignments. Source: Yu and Tajabji, 2007.

213 RIGID PAVEMENT BEST PRACTICES TABLE 3.12. DOWEL MISALIGNMENT AND EFFECTS ON PAVEMENT PERFORMANCE Type of Misalignment Effect on Spalling Slab Cracking Load Transfer Horizontal translation No No Yes Longitudinal translation No No Yes Vertical translation Yes No Yes Horizontal skew Yes Yes Yes Vertical tilt Yes Yes Yes Source: Federal Highway Administration, 2006. An illustration of a failed contraction joint due to dowel misalignment is shown in Figure 3.36. Additionally, an example of dowel “longitudinal translation” is also shown. A critical step for minimizing misalignment is to measure the postconstruction location of the dowel bars. There are multiple ways this can be done, but an instru- ment available from Magnetic Imaging Tools (MIT) is explored here. The device, MIT Scan-2, has been assessed and described by FHWA studies (Yu and Khazanovich, 2005; Yu, 2005) and applied on numerous paving projects. The nondestructive instrument uses magnetic tomography to locate metal objects (steel dowels for this application). This process is, in essence, an imaging technique that induces currents in steel dowels, and these currents provide the needed location information. A MIT Scan-2 device is shown in operation in Figure 3.37. The MIT Scan-2 has daily productivity rates of about 250 doweled joints for a single lane and can be used with freshly placed or hardened concrete. The FHWA, through its Concrete Pavement Technology Program (CPTP), has three of these units available to the states for loan or on-site demonstration (as of April 2011). (a) Top of slab for a removed joint (b) Figure 3.36. Dowel misalignment from an Interstate pavement. (a) Failed contraction joint due to dowel misalign- ment. (b) Example of dowel longitudinal translation (joint is not the same as the one in the accompanying photo). Photos: Kevin Littleton and Joe Mahoney.

214 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Various studies have been performed to examine the issue of what are allowable dowel misalignments. A best practices document is available from FHWA (Yu and Tayabji, 2007). Example Designs Table 3.13 summarizes a selection of unbonded concrete overlays of concrete pave- ments constructed in the United States since 1993. The information presented in the table was compiled from National Concrete Overlay Explorer [a database provided by the American Concrete Pavement Association (2010)]. The website currently contains only a representative sampling of projects across the United States, and so the number of concrete overlay projects viewable online is expected to increase over time. The common features for these unbonded concrete overlays in Table 3.13 include the following: • Slab thickness ranges from 9 to 12 in.; • Doweled joints are spaced mostly at 15 ft; • HMA interlayers range in thickness from 1 to 3 in. with most dense-graded but some open-graded mixes; and • Existing pavements were either jointed or CRCP. Figure 3.37. MIT Scan-2. Source: Yu and Khazanovich, 2005.

215 RIGID PAVEMENT BEST PRACTICES TABLE 3.13. A SELECTION OF UNBONDED CONCRETE OVERLAYS CONSTRUCTED IN THE UNITED STATES SINCE 1993 Project Location and Details Year of Overlay Construction Design Details of Overlay I-77, Yadkin, South of Elkin, North Carolina. The existing pavement is CRCP and 30 years old. 2008 • Slab thickness is 11 in. • Doweled joints spaced at 15 ft • Asphalt 1.5-in. interlayer I-86, Olean, New York. The existing pavement is JRCP and 30 years old. 2006 • Slab thickness is 9 in. • Doweled joints spaced at 15 ft • Asphalt 3-in. interlayer • 30% truck traffic I-35, Noble/Kay County, Oklahoma. The existing pavement is JRCP and 42 years old. 2005 • Slab thickness is 11.5 in. • Doweled joints spaced at 15 ft • Asphalt 2-in. interlayer • 25% truck traffic I-40, El Reno, Oklahoma. The existing pavement is JPCP and 35 years old. 2004 • Slab thickness is 11.5 in. • Doweled joints spaced at 15 ft • Asphalt 2-in. interlayer I-264, Louisville, Kentucky. The existing pavement is JRCP and 36 years old. 2004 • Slab thickness is 9 in. • Doweled joints spaced at 15 ft • Drainable asphalt 1-in. interlayer I-40, El Reno, Oklahoma (MP 119 and east). Existing pavement is JPCP and 34 years old. 2003 • Slab thickness is 10 in. • Doweled joints • Asphalt 2-in. interlayer I-85 (southbound), near Anderson, South Carolina. Existing pavement is JPCP and 38 years old. 2002 • Slab thickness is 12 in. • Doweled joints • Asphalt 2-in. interlayer • 35% truck traffic • The northbound lanes have been rubblized and overlaid. Performance comparison is recommended. I-275, Circle Freeway, Kentucky. Existing pavement is JPCP and 28 years old. 2002 • Slab thickness is 9 in. • Doweled joints spaced at 15 ft • Drainable asphalt 1-in. interlayer I-65, Jasper County, Indiana. Existing pavement is JRCP and 25 years old. 1993 • Slab thickness is 10.5 in. • Doweled joints spaced at 20 ft • Asphalt 1.5-in. interlayer • 23% truck traffic I-40, Jackson, Tennessee. Existing pavement is JPCP. 1997 • Slab thickness is 9 in. • Doweled joints spaced at 15 ft • Asphalt 1-in. interlayer I-85, Granville, North Carolina. Existing pavement is CRCP and 25 years old. 1998 • Slab thickness is 10 in. • Doweled joints spaced at 18 ft • Permeable asphalt 2-in. interlayer • 25% truck traffic continued

216 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Summary for Unbonded Concrete Overlays of Concrete Pavements Based on the review of the best practices and performance of pavement sections in the LTPP database and related data in these best practices, the design recommendations for long-lived unbonded concrete overlays are summarized in Table 3.14. A selection of significant practices and specifications associated with pav- ing unbonded concrete overlays over existing concrete were selected and included in Table 3.15. The table includes a brief explanation of why the issue is of special interest, along with examples from the recommendations in the Guide Specifications (Chapter 4). Three major practices are featured: (1) existing pavement and preoverlay repairs, (2) overlay thickness and joint details, and (3) interlayer requirements. Project Location and Details Year of Overlay Construction Design Details of Overlay I-265 at I-71, Jefferson County, Kentucky. Existing pavement is JRCP and was constructed in 1970. 1999 • Slab thickness is 9 in. • Doweled joints spaced at 15 ft • Drainable asphalt 1.3-in. interlayer I-85, Newman, Georgia. Existing pavement is JPCP and 38 years old. 2009 • Slab thickness is 11 in. • CRCP overlay • Asphalt 3-in. interlayer Source: American Concrete Pavement Association, 2010. TABLE 3.13. A SELECTION OF UNBONDED CONCRETE OVERLAYS CONSTRUCTED IN THE UNITED STATES SINCE 1993 (continued) TABLE 3.14. RECOMMENDED DESIGN ATTRIBUTES FOR LONG-LIFE CONCRETE PAVEMENT (LLCP) Design Attribute Recommended Range Slab thickness Minimum thickness of 9 in. Interlayer thickness ≥1 in.; 2 in. is likely optimal Joint spacing Maximum spacing of 15 ft Load-transfer device Mechanical load-transfer device, corrosion-resistant dowels to promote long-life Dowel lengths of 18 in. Dowel diameter 1.5 in. (function of slab thickness)

217 RIGID PAVEMENT BEST PRACTICES TABLE 3.15. SUMMARY OF BEST PRACTICES AND SPECIFICATIONS FOR UNBONDED CONCRETE OVERLAYS OVER EXISTING CONCRETE Best Practice Why This Practice? Typical Specification Requirements Existing pavement and preoverlay repairsa The preparation of the existing pavement is important for achieving long life from the unbonded concrete overlay. Existing Pavement Condition Possible Repairs Faulting ≤10 mm No repairs needed Faulting >10 mm Use a thicker interlayer Significant tenting, shattered slabs, pumping Full-depth repairs Severe joint spalling Clean the joints CRCP with punchouts Full-depth repairs Overlay thickness and joint detailsb Thickness and joint details are critical for long-life performance. • Overlay thickness ≥9 in. • Transverse joint spacing not to exceed 15 ft when slab thicknesses are in excess of 9 in. • Joints should be doweled; dowel diameter should be a function of slab thickness. The recommended dowel bar sizes are: — For ≥9 in.: 1.50-in.-diameter minimum • Dowels should be corrosion resistant Interlayer between overlay and existing pavementb Interlayer thickness and conditions prior to placing the concrete overlay influence long-life performance and early temperature stress in the new slabs. • The interlayer material shall be a minimum of 1-in.-thick new bituminous material. • The surface temperature of HMA interlayer shall be <90°F before overlay placement. Concrete overlay materialsb • Supplementary cementitious materials may be used to replace a maximum of 40%–50% of the portland cement. a For additional details, see Elements for AASHTO Specifications 552, 557, and 558 in Chapter 4. b For additional details, see Elements for AASHTO Specification 563 in Chapter 4. UNBONDED CONCRETE OVERLAYS OF HOT-MIX ASPHALT CONCRETE PAVEMENTS Criteria for Long-Life Potential Unbonded concrete overlays of HMA concrete pavements are a viable long-lived re- newal strategy. In general, this strategy is applied when the existing HMA pavements exhibit significant deterioration in the form of rutting, fatigue cracking, potholes, foundation issues, and pumping; however, the stability and the uniformity of the exist- ing pavement are important for both renewal construction and long-life performance of the unbonded concrete overlay. Figure 3.38 is a sketch of an unbonded overlay over preexisting flexible pavement.

218 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE The placement of the overlay can potentially do the following (Smith et al., 2002; Harrington, 2008): • Restore and/or enhance structural capacity of the pavement structure; • Increase life equivalent to that of a full-depth pavement; and • Restore and/or improve friction, noise, and rideability. General Design Considerations The structural condition of the existing pavement can be established by conducting visual distress surveys and deflection testing using an FWD. The deflection information can be used to back-calculate the resilient moduli of various pavement layers (although HMA layers less than 3 in. thick are difficult to back-calculate). Preoverlay Repairs The preoverlay requirements are minimal at best. Table 3.16 summarizes the possible preoverlay repairs needed in preparation for the PCC unbonded concrete overlay of asphalt pavements (Harrington, 2008). TABLE 3.16. SUGGESTED PREOVERLAY REPAIRS Existing Pavement Condition Possible Repairs Potholes Fill with asphalt concrete Shoving Mill Rutting ≥2 in. Mill Rutting <2 in. None or mill Crack width ≥4 in. Fill with asphalt Source: Harrington, 2008. Figure 3.38. Unbonded concrete overlay of flexible pavement. Illustration: Joe Mahoney.

219 RIGID PAVEMENT BEST PRACTICES Structural Design The design of an unbonded concrete overlay of HMA pavement considers the existing pavement as a stable and uniform base, and the overlay thickness is designed similarly to a new concrete pavement. Furthermore, the design assumes an unbonded condi- tion between the existing asphalt layer and the new concrete overlay. The exist ing asphalt thickness should be at least 4 in. of competent material to ensure an adequate load- carrying base for the concrete overlay (Smith et al., 2002; Harrington, 2008). The 1993 AASHTO design method does not consider the effects of bonding between the new overlay and the existing HMA pavement. The design method considers the composite k at the top of the HMA layer. Field studies have shown that there is some degree of bonding between the two layers. However, the longevity and the uniformity of this bond over the design life of the structure is not well documented. In the MEPDG design procedure, the bonding between the two layers is modeled by selecting appro- priate friction factors. In general (as documented in the literature), the unbonded overlay thickness usu- ally ranges from 4 to 11 in.; however, to ensure long-life performance the slab thick- nesses of the overlay should range from 9 to 13 in. The joint design, slab length, and joint width details are similar to unbonded concrete overlays of concrete pavements. Performance Considerations In general, the field performance of unbonded concrete overlays of HMA pavements has been satisfactory. The success of the renewal strategy hinges on the uniform under- lying support. The underlying HMA base eliminates most of the pumping of fines so there is little to no faulting and very uniform support. The general performance of PCC over HMA has been very good. Example Designs Table 3.17 summarizes unbonded concrete overlays of concrete pavements constructed in the United States since 1995. The information presented in the table was compiled from the National Concrete Overlay Explorer (American Concrete Pavement Associa- tion, 2010). The website currently contains only a representative sampling of projects across the United States, and so the number of concrete overlay projects viewable online is expected to increase over time. The common features for these unbonded concrete overlays in Table 3.17 include the following: • Slab thicknesses range from 9 to 12 in., and • Doweled joints are spaced mostly at 15 ft.

220 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 3.17. OVERVIEW OF SELECTED UNBONDED CONCRETE OVERLAYS OF FLEXIBLE PAVEMENTS CONSTRUCTED IN THE UNITED STATES SINCE 1995 Project Location and Details Year of Overlay Construction Design Details of Overlay Cherry Street, North to H-17, Iowa. 2004 • Slab thickness is 9 in. • Doweled joints spaced at 15 ft Tiger Mountain, Oklahoma. Existing pavement was 9 years old. 2004 • Slab thickness is 10.5 in. • Doweled joints spaced at 15 ft • 30% truck traffic US-412, Bakerville, Missouri. Existing pavement is 30 years old. 2004 • Slab thickness is 12 in. • Doweled joints spaced at 15 ft • 24% truck traffic US-412, Bakerville, Missouri. 2003 • Slab thickness is 12 in. • Doweled joints spaced at 15 ft • 24% truck traffic I-55, Vaiden, Mississippi. 2001 • Slab thickness is 10 in. • Doweled joints spaced at 16 ft E-33, Iowa. 1998 • Slab thickness is 9 in. • Doweled joints spaced at 15 ft P-33, Iowa. 1998 • Slab thickness is 10 in. • Doweled joints spaced at 15 ft I-10/1-12, Louisiana. 1995 • Slab thickness is 12 in. Source: Data from American Concrete Pavement Association, 2010. ADDED LANES AND TRANSITIONS FOR ADJACENT STRUCTURES FOR UNBONDED PCC OVERLAYS OVER EXISTING CONCRETE AND HMA PAVEMENTS There is little guidance found in the literature on integrating new or rehabilitated pave- ments into adjacent pavements and features. This document addresses adding lanes to an existing pavement structure, as well as accommodating existing features such as bridge abutments and vertical clearance restrictions within the limits of a pave- ment renewal project. These issues are paramount when using the existing pavement in place as part of long-life renewal, because there is typically a significant elevation change associated with each renewal alternative. The following recommendations are based on discussions with the state highway agencies surveyed in Phase 1 and those agencies that participated in Phase 2. Bridge and Overcrossing Structure Approaches In the transition where the unbonded PCC overlay connects to a bridge approach, or when the roadway section with an unbonded overlay passes under an existing struc- ture, the new grade line and reduced vertical clearances usually require the construction

221 RIGID PAVEMENT BEST PRACTICES of a new pavement section. The length of the new section depends on the elevation difference, but it is usually in the range of 300 to 500 ft before and after the structure. A typical taper rate used by a number of agencies visited is 400 to 1 to transition from the new grade line to the elevation required by the adjacent feature. Attention should be paid to the longitudinal drainage as well as to the transverse drainage when designing the new pavement section. Where possible, the existing subgrade elevation and grade should be maintained in the longitudinal direction as well as the transverse direction. Because the new roadway section will not be as thick as the renewal approach using the existing pavement, the difference in elevation is usually made up with HMA or a combination of HMA and untreated granular base material. Because the unbonded PCC overlay requires reasonably uniform support, the transition from the old PCC pavement to the new pavement should be made as stiff as possible, which may require replacement of the PCC with full-depth HMA. Subgrade stabilization should also be considered if needed in the transition area. Specifically, the SHRP 2 Renewal Project R02 guidance for “Geotechnical Solutions for Transportation Infrastructure” and its recommendations for stabilization of the pavement working platform should be con- sidered. Diagrams of possible transition profiles are shown in Figures 3.39 and 3.40. Figure 3.40. Diagram of transition beneath structure. Figure 3.39. Diagram of transition to bridge approach (unbonded PCC overlay of PCC pavement).

222 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE In some cases, agencies reported they were able to raise an overcrossing rather than reconstruct the roadway for less cost and reduced impact on traffic. That option may be considered where possible, particularly in more rural areas where there is little cross traffic on the overcrossing. Added Lanes or Widening When a project calls for additional lanes or widening, the addition of lanes often facili- tates the staging of the traffic through the project, but it usually produces a mismatch in pavement sections in the transverse direction. The slope and grade line of the sub- grade should be maintained so that water flowing along the contact between the base and the subgrade does not get trapped in the transverse direction. There is a risk that there may be reflection cracking between the existing pavement and the new pavement section, particularly when the existing pavement is a PCC. Also of concern is the need for stabilizing the subgrade soil, if required for widening. Subgrade stabilization will increase the stability of the roadway section, accelerate pavement construction, and help reduce some of the settlement or differential vertical deflection that causes reflec- tion cracking along the contact with the old PCC pavement. Specifically, the SHRP 2 Renewal Project R02 guidance for “Geotechnical Solutions for Transportation Infra- structure” and its recommendations for stabilization of the pavement working plat- form should be considered. Lane Widening A number of agencies have reported they have constructed a 14-ft widened lane in the outside lane to provide improved edge support. One agency reported cracking along the edge of the old PCC pavement caused by nonuniform support at that location. They had not improved the shoulder section prior to construction of the unbonded PCC overlay. If lane widening is considered, the existing shoulder section may need to be reconstructed to provide more uniform support for the new PCC pavement. Added Lanes When a project calls for additional lanes or widening, the addition of lanes often facili- tates the staging of the traffic through the project, but it usually produces a mismatch in pavement sections in the transverse direction. The slope and grade line of the sub- grade should be maintained so that water flowing along the contact between the base and the subgrade does not get trapped in the transverse direction. Similar to widened lanes, there is a need for uniform support under the PCC overlay; thus, the shoulder will need to be reconstructed and the subgrade should be stabilized where needed. No specific guidance could be found to provide uniform support in the widening next to the existing PCC pavement. A number of agencies have widened with HMA as part of the traffic staging and then placed the unbonded PCC pavement across both the existing PCC pavement with a HMA bond breaker, and the widened HMA pave- ment. Some agencies have widened the existing PCC pavement with PCC pavement, then placed the HMA bond breaker across both the old and the new PCC pavement before placing the PCC overlay. This approach provides uniform support for the PCC overlay; however, there was no indication that there was any difference in performance

223 RIGID PAVEMENT BEST PRACTICES when the widening was constructed with PCC pavement or HMA pavement as a base for the PCC overlay. Use of HMA to widen the existing pavement does provide some advantage in traffic staging. Typical pavement sections are shown in Figures 3.41 and 3.42. The minimum thickness of the HMA in the widening is usually controlled by the traffic loading during staging, but it is usually a minimum of 6 in. thick to minimize failure risk during staging and provide more uniform support for the PCC overlay. For unbonded PCC overlays of flexible pavement, the existing pavement is simply widened with HMA to provide the base for the PCC overlay. The pavement section should extend the subgrade line and slope out to either the contact with the in-slope of the ditch or fill slope, or to a collection point for longitudinal drains as shown in Figures 3.42 and 3.43. Figure 3.41. Cross section showing existing PCC pavement without daylighted shoulders. Figure 3.42. Cross section showing widening of the shoulder with daylighting or drainage.

224 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE BEST PRACTICES SUMMARY The definition of long-life renewal strategies is a design life of 50 years or more. To achieve this, unbonded concrete overlays of existing pavements are recommended. This recommendation is based on several sets of information, including but not limited to (1) state DOT criteria, (2) LTPP findings, and (3) information from the National Concrete Pavement Technology Center. To achieve a 50-year life, several practices are critical, and these include the selection of materials, knowledge of local pavement distress and its causes, struc- tural design, and relevant construction practices. Two broad types of unbonded con- crete were discussed: (1) unbonded concrete over existing concrete pavement and (2) unbonded concrete over existing HMA pavement. Concrete overlays can be either JPCP or CRCP—both perform well. Table 3.15 is a summary of relevant best practices and related specification require- ments for unbonded concrete overlays. Three major practices are featured: (1) existing pavement and preoverlay repairs, (2) overlay thickness and joint details, and (3) inter- layer requirements. The major findings are recapped in Table 3.18. Figure 3.43. Cross-section detail with PCC shoulder.

225 RIGID PAVEMENT BEST PRACTICES REFERENCES AASHTO. “AASHTO Guide for Design of Pavement Structures,” American Association of State Highway and Transportation Officials, Washington, D.C., 1993. AASHTO. “Provisional Practice for Determining the Reactivity of Concrete Aggregates and Selecting Measures for Preventing Deleterious Expansion in New Concrete Construction,” PP 65-10, Provisional Standards, 14th ed., American Association of State Highway and Transportation Officials, Washington, D.C., 2010. American Concrete Pavement Association. “Guidelines for Unbonded Concrete Overlays,” Technical Bulletin 005.0D, American Concrete Pavement Association, Skokie, Ill., 1990. American Concrete Pavement Association. “Pavement Surface Characteristics—A Synthesis and Guide,” American Concrete Pavement Association, Skokie, Ill., 2006. American Concrete Pavement Association. “The National Concrete Overlay Explorer,” American Concrete Pavement Association, Skokie, Ill., 2010. http://overlays.acpa.org/ webapps/overlayexplorer/index.html. Anderson, K., J. Uhlmeyer, L. Pierce, and J. Weston. “Wear Resistant Pavement Study,” Report WA-RD 657.1, Washington State Department of Transportation, Olympia, 2006. Applied Pavement Technology. “Unbonded Portland Cement Concrete Overlays,” Technical Brief FHWA-IF-03-006, Federal Highway Administration, 2002. ARA, Inc. “Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures,” Final Report, Part 2: Design Inputs, Chapter 2: Material Characterization, National Cooperative Highway Research Program, Transportation Research Board of the National Academies, Washington, D.C., 2004. TABLE 3.18. SUMMARY OF RECOMMENDED PRACTICES FOR UNBONDED PCC OVERLAYS Factor or Consideration Practice Concrete overlay thickness ≥9 in. Type of concrete overlay Unbonded JPCP or CRCP Structural design Do a complete structural design using an agency-approved method. JPCP joint spacing ≤15 ft JPCP load transfer Use 1.5-in.-diameter dowel bars. Type of dowel bar Use corrosion-resistant dowels. Aggregates Use local state DOT specifications with special attention paid to eliminating the potential for ASR and D-cracking. Cements SCM is acceptable and may be superior to traditional portland cements; use state guidelines for max limits. Existing pavement Use criteria provided for preoverlay repairs. Concrete overlay interlayer Use an HMA interlayer 1 (minimum) to 2 in. thick. Concrete overlay construction Control mix and substrate temperatures during construction; tools such as HIPERPAV will help planning and execution.

226 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Burnham, T. “Thin and Ultra-Thin Concrete Overlays,” MnRoad Lessons Learned, Office of Materials, Minnesota Department of Transportation, 2008. Concrete Pavement Technology Center. “Guide to Dowel Load Transfer Systems for Jointed Concrete Pavements,” National Concrete Consortium, National Concrete Pavement Tech- nology Center, Iowa State University, Ames, 2011. Environmental Protection Agency. “Resource Conservation—Comprehensive Procurement Guidelines,” Environmental Protection Agency, 2011. http://www.epa.gov/epawaste/con- serve/tools/cpg/about.htm. ERES Consultants, Inc. “Evaluation of Unbonded Portland Cement Concrete Overlays,” NCHRP Report 415, Transportation Research Board, Washington, D.C., 1999. Farney, J., and S. Kosmatka. “Diagnosis and Control of Alkali-Aggregate Reactions in Concrete,” IS 413, Concrete Information, Portland Cement Association, Skokie, Ill., 1997. Federal Highway Administration. “High Performance Concrete Pavements—Project Summary,” Publication FHWA-IF-06-031, Federal Highway Administration, 2006. Folliard, K., and K. Smith. “Aggregate Tests for Portland Cement Concrete Pavements: Review and Recommendations,” NCHRP Research Results Digest, No. 281, National Cooperative Highway Research Program, Transportation Research Board of the National Academies, Washington, D.C., 2003. Fournier, B., M. Berube, K. Folliard, and M. Thomas. “Report on the Diagnosis, Prognosis, and Mitigation of Alkali-Silica Reaction (ASR) in Transportation Structures,” Report FH- WA-HIF-09-001, Office of Pavement Technology, Federal Highway Administration, 2010. Germann Instruments, Inc. “Eddy-Dowel.” www.germann.org/TestSystems/Eddy-Dowel/ Eddy-Dowel.pdf. Accessed July 2011. Harrington, D. “Guide to Concrete Overlays: Sustainable Solution for Resurfacing and Rehabilitating Existing Pavements,” 2nd ed., National Concrete Pavement Technology Center, Ames, Iowa, 2008. HIPERPAV. “HIPERPAV High Performance Paving Software,” Version 3.20.0006, Transtec Group and the Federal Highway Administration, 2010. http://www.hiperpav.com/. Kim, D. H., S. C. Choi, Y. H. Cho, and M. C. Won. “Long-Term Performance of Thin Bonded Concrete Overlay in Texas,” Report 07-0031. Presented at 86th Annual Meeting of the Transportation Research Board, Washington, D.C., 2007. McGhee, K. “NCHRP Synthesis of Highway Practice 204: Portland Cement Concrete Resurfacing,” TRB, National Research Council, Washington, D.C., 1994. Miner, M. “Cumulative Damage in Fatigue,” Transactions of the American Society of Mechanical Engineers, Vol. 67, 1945. Minnesota Department of Transportation. “Standard Specifications for Road, Bridge, and Municipal Construction,” Minnesota Department of Transportation, St. Paul, 2005a. Minnesota Department of Transportation. “S114 – (2301) High Performance Concrete Pavement—Dowel Bar,” Special Provisions, Minnesota Department of Transportation, St. Paul, 2005b. Missouri Department of Transportation. “SHRP LTPP Test Sections,” Research Investiga- tion 91-001, Research, Development, and Technology Division, Missouri Department of Transportation, Jefferson City, 1998.

227 RIGID PAVEMENT BEST PRACTICES Missouri Department of Transportation. “Missouri Guide for Pavement Rehabilitation,” Report RDT 02-013, Research, Development, and Technology Division, Missouri Depart- ment of Transportation, Jefferson City, 2002. MnRoad. “Whitetopping: Concrete Overlays of Asphalt Pavements,” Version 1, Minnesota Department of Transportation, St. Paul, 2009. Neville, A. Properties of Concrete, John Wiley and Sons, New York, 1975. Packard, R. “Design of Concrete Airport Pavement,” Engineering Bulletin, Portland Cement Association, Skokie, Ill., 1973. Packard, R. “Thickness Design for Concrete Highway and Street Pavements,” Engineering Bulletin, Portland Cement Association, 1984. Rasmussen, R., and D. Rozycki. “NCHRP Synthesis 338: Thin and Ultra-Thin Whitetopping—A Synthesis of Highway Practice,” TRB, National Research Council, Washington, D.C., 2004. Rasmussen, R., S. Garber, G. Fick, T. Ferragut, and P. Wiegard. “How to Reduce Tire- Pavement Noise: Interim Better Practices for Constructing and Texturing Concrete Pave- ment Surfaces,” Pooled Fund TPF-5(139) PCC Surface Characteristics: Tire-Pavement Noise Program Part 3—Innovative Solution/Current Practices, Federal Highway Administration, 2008. Richardson, D. “Aggregate Gradation and Optimization—Literature Search,” Report RDT 05-001, Missouri Department of Transportation, Jefferson City, 2005. Smith, K., H. Yu, and D. Peshkin. “Portland Cement Concrete Overlays: State of the Tech- nology Synthesis,” Report FHWA-IF-02-045, Federal Highway Administration, 2002. Smith, K. D., H. T. Yu, M. J. Wade, D. G. Peshkin, M. I. Darter. “Performance of Concrete Pavements. Volume 1: Field Investigation,” Report FHWA-RD-94-177. Federal Highway Administration, 1997. Smith, T., and S. Tayabji. “Assessment of the SPS-7 Bonded Concrete Overlays Experi- ment,” Final Report, FHWA-RD-98-130, Office of Engineering Research and Development, Federal Highway Administration, 1998. Snyder, M. B., M. J. Reiter, K. T. Hall, and M. I. Darter. “Rehabilitation of Concrete Pave- ments. Volume 1: Repair Rehabilitation Techniques,” Report FHWA-RD-88-071, Federal Highway Administration, 1989. Tayabji, S., and S. Lim. “Long Life Concrete Pavements: Best Practices and Directions from the States,” Concrete Pavement Technology Program, Tech Brief, FHWA-HIF-07-030, Federal Highway Administration, 2007. Tayabji, S., A. Gisi, J. Blomberg, and D. DeGraaf. “New Applications for Thin Concrete Overlays: Three Case Studies,” National Conference on PRR of Concrete Pavements, St. Louis, Mo., 2009. Texas Department of Transportation. Chapter 8: Thin Concrete Overlays, and Table 8-4: Thin Whitetopping Thickness Design, “Pavement Design Manual,” Texas Department of Transportation, Austin, 2008. Texas Department of Transportation. “Continuously Reinforced Concrete Pavement—One Layer Steel Bar Placement,” Design Division Standard, Texas Department of Transporta- tion, Austin, 2009a.

228 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Texas Department of Transportation. “Continuously Reinforced Concrete Pavement—Two Layer Steel Bar Placement,” Design Division Standard, Texas Department of Transporta- tion, Austin, 2009b. Texas Department of Transportation. Chapter 8, Section 5: Determining Concrete Pavement Thickness, “Pavement Design Guide,” Texas Department of Transportation, Austin, 2011. Thomas, M., B. Fournier, and K. Folliard. “Report on Determining the Reactivity of Con- crete Aggregates and Selecting Appropriate Measures for Preventing Deleterious Expansion in New Concrete Construction,” Report FHWA-HIF-09-001, Office of Pavement Technol- ogy, Federal Highway Administration, 2008. Van Dam, T., L. Sutter, D. Smith, M. Wade, and K. R. Peterson. “Guidelines for Detection, Analysis, and Treatment of Material Related Distresses in Concrete Pavements,” Volume 1: Final Report, FHWA-RD-01-163, Federal Highway Administration, 2002. Washington State Department of Transportation. “Standard Specifications for Road, Bridge, and Municipal Construction,” M41-10, Washington State Department of Transportation, Olympia, 2010. Watson, M., E. Lukanen, S. Olson, and T. Burnham. “Construction Report for a Thin Unbonded Concrete Overlay on Minnesota TH 53,” Interim Report 2010-23, Minnesota Department of Transportation, St. Paul, 2010. Wisconsin Department of Transportation. “HPC Pavement Requirements for Concrete Pavement 12-inch,” Special Provisions, Wisconsin Department of Transportation, Madison, 2009. Yu, H. T. “Use of Magnetic Tomography Technology to Evaluate Dowel Bar Placement,” Report FHWA-IF-06-002, Federal Highway Administration, 2005. Yu, H. T., and L. Khazanovich. “Use of Magnetic Tomography Technology to Evaluate Dowel Placement,” Report FHWA-IF-06-006, Federal Highway Administration, 2005. Yu, H. T., and S. Tayabji. “Best Practices for Dowel Placement Tolerances,” Report FHWA- HIF-07-021, Federal Highway Administration, Washington, D.C., 2007.