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.

1INTRODUCTION Why This Assessment Manual? This assessment manual was prepared to aid the process of renewing existing pave- ments so that long lives can be achieved. To achieve this goal, a systematic collection of relevant pavement-related data is needed. Further, such data need to be organized to maximize the usefulness in the pavement decision-making process. To that end, this manual will help. The types of data collection contained in this manual range from basic information such as distress surveys to insights on traffi c impacts. The last section provides informa- tion on life-cycle assessments (environmental accounting). The use of this type of assess- ment is receiving increased attention and is likely to be widely applied in the future. How to Use the Manual The manual is intended to complement the design tools developed by SHRP 2’s R23 study. The types of data critical for making pavement-related decisions are described along with methods (analysis tools) for using the information in decision-making ap- plications. It is not assumed that all data categories will be collected or assessed for a specifi c renewal project. Rather, the manual is designed as a reference document that provides information relevant to all renewal strategies considered in SHRP 2 Renewal Project R23. Assessment Data Categories The following 11 categories are described in this manual: • Pavement distress survey, • Pavement rut depth and roughness, 1 PROJECT ASSESSMENT MANUAL

2GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE • Nondestructive testing via the falling weight deflectometer (FWD), • Ground-penetrating radar (GPR), • Pavement cores, • Dynamic cone penetrometer (DCP), • Subgrade soil sampling and tests, • Traffic loads for design, • Construction productivity and traffic impacts, • Life-cycle assessment (environmental accounting), and • Miscellaneous material properties. Each data category is structured much the same way, into (1) the purpose of collecting the data, (2) applicable standards, definitions, and data organization recommenda- tions, and (3) analysis tools. Overall Assessment Scheme The overall assessment scheme performed by the user can range from rather basic in- formation about the existing and proposed pavement structure to substantially more detailed data and analyses. The basic scheme is illustrated in Figure 1.1. Figure 1.1. Outline of assessment scheme.

3PROJECT ASSESSMENT MANUAL The first three boxes (1 through 3) shown in Figure 1.1 are addressed in this assess- ment manual, with that information being applied to the processes shown in the last two boxes (4 and 5). PAVEMENT DISTRESS SURVEY Purpose This section provides an overview of the use of a pavement distress survey to help make pavement assessment decisions. Measurement Methods This subsection is used to describe definitions and standards applicable to pavement distresses and provides a way to organize such information. Pavement Distress Measurements ASTM D6433-07: Standard Practice for Roads and Parking Lots Pavement Condition Index Surveys. Distress Identification Manual for the Long-Term Pavement Performance Program FHWA-RD-03-031, June 2003. Discussion Pavement distress data can be used for numerous purposes, but three are noted: (1) establish pavement reconstruction, rehabilitation, and maintenance priorities, (2) determine rehabilitation and maintenance strategies, and (3) predict pavement per- formance. This type of information is a key element for decision making associated with pavement renewal options. McCullough (1971) provided a detailed description of three basic pavement dis- tress groups, associated modes, and examples as shown in Table 1.1. The majority of distress survey protocols use a subset of fracture, distortion, and/or disintegration. Upon closer inspection of Table 1.1 for flexible pavements, two of these—fracture and disintegration—cause most pavement rehabilitation and maintenance actions. More specifically, these can be categorized by fatigue, transverse cracking, and strip- ping or raveling. Tables 1.2, 1.3, and 1.4 provide templates for flexible pavement dis- tress data collection. It is assumed that cores will be an integral part of the pavement distress examination; hence, locations would logically be organized by mileposts or another appropriate location referencing system. For multilane highways, this infor- mation can be collected for the design lane or all lanes in one direction, as per project requirements.

4GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 1.1. DISTRESS GROUPS Distress Group Distress Mode Examples of Distress Mechanism Fracture Cracking Excessive loading Repeated loading (i.e., fatigue) Thermal changes Moisture changes Slippage (horizontal forces) Shrinkage Spalling Excessive loading Repeated loading (i.e., fatigue) Thermal changes Moisture changes Distortion Permanent deformation Excessive loading Time-dependent deformation (e.g., creep) Densification (i.e., compaction) Consolidation Swelling Frost Faulting Excessive loading Densification (i.e., compaction) Consolidation Swelling Disintegration Stripping Adhesion (i.e., loss of bond) Chemical reactivity Abrasion by traffic Raveling and scaling Adhesion (i.e., loss of bond) Chemical reactivity Abrasion by traffic Degradation of aggregate Durability of binder Source: After McCullough, 1971. The following distress types should be measured and recorded if present on the existing pavement: • Flexible pavement distress (definitions from or modified after LTPP Distress Man- ual, Miller and Bellinger, 2003) 1. Fatigue cracking occurs in areas subjected to repeated traffic loadings (wheel- paths). It can be a series of interconnected cracks in early stages of development and develops into many-sided, sharp-angled pieces, usually less than 0.3 m on the longest side, characteristically with a chicken wire/alligator pattern in later stages. For illustrations of fatigue-cracking severity levels, see Figure 1.2.

5PROJECT ASSESSMENT MANUAL 2. Transverse cracking involves cracks that are predominantly perpendicular to the pavement centerline. For illustrations of transverse-cracking severity levels, see Figure 1.3. 3. Stripping or raveling is the wearing away of the pavement surface caused by the dislodging of aggregate particles and loss of asphalt binder. Raveling ranges from loss of fines to loss of some coarse aggregate and ultimately to a very rough and pitted surface with obvious loss of aggregate (e.g., see Figure 1.4). This study expands the definition to identification of stripping or raveling in the surface layer to include stripping that may be occurring in lower hot-mix asphalt (HMA) layers in the pavement structure. The depth of stripping can be verified by GPR analyses and/or coring as discussed in the sections “Ground- Penetrating Radar (GPR)” and “Pavement Cores” in this chapter. • Rigid pavement distress for jointed plain concrete pavement (JPCP), jointed rein- forced concrete pavement (JRCP), and continuously reinforced concrete pavement (CRCP) (definitions from or modified after LTPP Distress Manual, Miller and Bellinger, 2003, with the exception of alkali-silica reactivity (ASR) cracking) 1. Pavement cracking includes all major types of cracks that can occur in a slab. This can include corner breaks, and longitudinal and transverse cracking as de- fined by Miller and Bellinger (2003). Corner break cracks intersect the adjacent transverse and longitudinal joints at approximately a 45° angle. Longitudinal cracking and transverse cracking are parallel and transverse to the centerline, respectively. For an example of a portland cement concrete (PCC) slab with multiple cracks, see Figure 1.5. 2. Joint faulting is the difference in elevation across a joint or crack. For an ex- ample of joint faulting, see Figure 1.6. 3. Materials-caused distress includes (1) D-cracking, a closely spaced crescent- shaped hairline cracking pattern that occurs adjacent to joints, cracks, or free edges, with a dark coloring of the cracking pattern and surrounding area, some- times referred to as durability cracking (see Figure 1.7); and (2) ASR cracking, which is cracking of the PCC that can be easily confused with D-cracking or shrinkage cracking (see Figure 1.8). AASHTO has issued a Provisional Practice (AASHTO Designation PP 65-10) to address ASR. For complete details, please reference Chapter 3 of this Guide, “Rigid Pavements Best Practices.” 4. Pumping is the ejection of water from beneath the pavement. In some cases, detectable deposits of fine material are left on the pavement surface, which were eroded (pumped) from the support layers and have stained the surface. 5. Punchouts are composed of the area enclosed by two closely spaced (usually <0.6 m) transverse cracks, a short longitudinal crack, and the edge of the pave- ment or a longitudinal joint. Punchouts also include “Y” cracks that exhibit spalling, breakup, or faulting.

6GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Pavement Distress Data Templates The templates for specific pavement distress types follow. TABLE 1.2. TEMPLATE FOR FLEXIBLE PAVEMENT DISTRESS: FATIGUE CRACKING Location (milepost) Depth Distress HMA (in.) Base (in.) Fatigue Cracking Severitya Extentb Depth of Fatigue Cracks (measured from the pavement surface)c Low Moderate High a Severity of fatigue cracking is low, medium, or high: (1) low = none or only a few connecting cracks, cracks are not spalled or sealed, and pumping not evident; (2) moderate = interconnected cracks forming a complete pattern, cracks may be slightly spalled, cracks may be sealed, and pumping is not evident; and (3) high = moderately or severely spalled interconnected cracks forming a complete pattern; pieces may move when subjected to traffic, cracks may be sealed, and pumping may be evident. The severity definitions are from the LTPP Distress Identification Manual (Miller and Bellinger, 2003). b Extent of fatigue cracking is based on percent of wheelpath areas. Record extent for each level of severity. c Depth of fatigue cracks can be full-depth or top-down cracking. This should be determined by the use of pavement cores. (a) (b) (c) Figure 1.2. Illustrations of fatigue-cracking severity levels. (a) Low severity. (b) Moderate severity. (c) High severity. Sources: (a) Pavement Interactive. (b) N. Jackson. (c) Pavement Interactive.

7PROJECT ASSESSMENT MANUAL TABLE 1.3. TEMPLATE FOR FLEXIBLE PAVEMENT DISTRESS: TRANSVERSE CRACKING Location (milepost) Depth Distress HMA (in.) Base (in.) Transverse Cracking Severitya Extentb Depth of Transverse Cracks (measured from the pavement surface)c Low Moderate High a Severity of transverse cracking is low, medium, or high: (1) low = unsealed cracks with a mean width ≤ 6 mm, sealed cracks with sealant material in good condition and with a width that cannot be determined; (2) moderate = cracks with mean widths > 6 mm and ≤ 19 mm, or any cracks with a mean width ≤ 19 mm and adjacent low-severity random cracking; and (3) high = cracks with a mean width of > 19 mm, or cracks with a mean width ≤ 19 mm and adjacent to moderate- to high-severity random cracking. The severity definitions are from the LTPP Distress Identification Manual (Miller and Bellinger, 2003). b Extent of transverse cracking is based on the number of cracks per 100 ft. Record extent for each level of severity. c Depth of fatigue cracks might be the full depth of the HMA or top-down cracking. This can only be determined by the use of pavement cores. (a) (b) (c) Figure 1.3. Illustrations of transverse-cracking severity levels. (a) Moderate severity. (b) Moderate to high s everity. (c) High severity. Sources: (a) Pavement Interactive. (b) WSDOT. (c) Pavement Interactive.

8GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 1.4. TEMPLATE FOR FLEXIBLE PAVEMENT DISTRESS: STRIPPING OR RAVELING Location (milepost) Depth Distress HMA (in.) Base (in.) Stripping or Raveling Extent (% of surface area) Full-depth stripping or raveling or confined to the wearing surface only? Observation must be based on cores. Note: Severity levels are not applicable for stripping; either it exists or it does not. Coring and/or GPR should be used to verify subsurface moisture damage. Figure 1.4. Illustration of raveling. Source: WSDOT Using Table 1.1 again, the most important JPCP distress types that initiate PCC pavement renewal actions are fracture (slab or pavement cracking), distortion (fault- ing, typically at transverse contraction joints), and disintegration, which includes mate- rials-caused distresses of D-cracking and ASR cracking. These are shown in Tables 1.5 through 1.8. Tables 1.9 and 1.10 apply to CRCP and composite pavements.

9PROJECT ASSESSMENT MANUAL TABLE 1.5. TEMPLATE FOR RIGID PAVEMENT DISTRESS (JPCP OR JRCP): PAVEMENT CRACKING Location (milepost) Depth Distress PCC Slab (in.) Base Pavement or Slab Cracking Typea Thickness (in.) % Slabs with Multiple Cracksb Comments a Three types of base underlying PCC: (1) granular base, (2) cement-treated base, and (3) asphalt-treated base. b Percentage of slabs with two or more pavement cracks. Figure 1.5. Examples of PCC slabs with multiple cracks. Sources: Pavement Interactive and J. Mahoney. TABLE 1.6. TEMPLATE FOR RIGID PAVEMENT DISTRESS (JPCP OR JRCP): FAULTING Location (milepost) Depth Distress PCC Slab (in.) Base Faulting Typea Thickness (in.) Average Fault Depth (in.) Comments a Three types of base underlying PCC: (1) granular base, (2) cement-treated base, and (3) asphalt-treated base. (a) (b) Figure 1.6. Examples of various levels of joint faulting. (a) Average fault ∼0.25–0.5 in. (b) Average fault ∼0.5 in. Source: Pavement Interactive.

10 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 1.7. TEMPLATE FOR RIGID PAVEMENT DISTRESS: D-CRACKING Location (milepost) Depth Distress PCC Slab (in.) Base D-Cracking Typea Thickness (in.) Severityb Extentc Comments Low Moderatee High a Three types of base underlying PCC: (1) granular base, (2) cement-treated base, and (3) asphalt-treated base. b Severity of D-cracking is low, medium (moderate), or high: (1) low = D-cracks are tight, with no loose or missing pieces, and no patching is in the affected area; (2) moderate = D-cracks are well defined, and some small pieces are loose or have been displaced; and (3) high = D-cracking has a well-developed pattern, with a significant amount of loose or missing material. Displaced pieces, up to 0.1 m2, may have been patched. c Extent is based on the amount of cracks or joints that exhibit D-cracking. This definition of extent is different than used by the LTPP report (Miller and Bellinger, 2003). Figure 1.7. Examples of D-cracking severity levels. (a) Low severity. (b) Low severity. (c) High severity. Sources: (a) Pavement Interactive and C. L. Monismith. (b) and (c) N. Jackson. (a) (b) (c) TABLE 1.8. TEMPLATE FOR RIGID PAVEMENT DISTRESS: ASR CRACKING Location (milepost) Depth Distress PCC Slab (in.) Base ASR-Related Cracking Typea Thickness (in.) Does ASR cracking apply to this pavement? Yes or Nob How was ASR detected or measured? a Three types of base underlying PCC: (1) granular base, (2) cement-treated base, and (3) asphalt-treated base. b Severity levels are not applicable for ASR. Either it exists or it does not.

11 PROJECT ASSESSMENT MANUAL Table 1.9 applies to CRCP. A critical distress for CRCP is punchouts (which falls under “fracture” in Table 1.1). For an illustration of a CRCP punchout, see Figure 1.9. (a) (b) Figure 1.8. Illustrations of ASR-cracking severity levels. (a) Early stage of cracking. (b) Advanced stage of cracking. Source: N. Jackson. TABLE 1.9. TEMPLATE FOR RIGID PAVEMENT DISTRESS (CRCP): PUNCHOUTS Location (milepost) Depth Distress PCC Slab (in) Base Punchouts Typea Thick (in) No./mile Comments a Three types of base underlying PCC: (1) granular base, (2) cement-treated base, and (3) asphalt-treated base. Figure 1.9. Advanced stage for a CRCP punchout. Source: FHWA.

12 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Other PCC pavement distress types can be important and such information can be collected and used; however, the distress types in the preceding tables were judged the most critical for pavement renewal decision making. Drainage Conditions An assessment of the existing pavement’s subsurface drainage is important in making pavement renewal decisions. The following factors, if observed, suggest that subsur- face drainage may be an issue and corrective actions may be needed for the renewal design process: • Pumping, • PCC joint or crack faulting, • Standing water in shallow ditches, and • Use of cement-stabilized base under PCC. Analysis Tools How pavement distress data are specifically used in the renewal decision-making pro- cess is covered in Appendices C and D of the R23 report. PAVEMENT RUT DEPTH AND ROUGHNESS Purpose This section overviews the use of pavement rut depths and roughness for aiding pave- ment assessment decisions. Measurement Methods This subsection is used to describe definitions and standards applicable for pavement rut and roughness measurements. TABLE 1.10. COMPOSITE PAVEMENT DISTRESSa Location (milepost) Depth Distressd HMA Surfacing (in.) PCC Describe Condition of Surface Course Comments PCC Typeb PCC Slab Thickness (in.) Base Typec Base Thickness (in.) Poor condition Very poor condition a Composite pavement definition assumes a flexible (HMA) layer overlies PCC. b Three types of PCC pavement: (1) JPCP, (2) JRCP, and (3) CRCP. c Three types of base underlying PCC: (1) granular base, (2) cement-treated base, and (3) asphalt-treated base. d Distress is broadly defined for composite pavements. The only initial information available to the user is the surface condition, which can include a range of distress types—most likely cracking.

13 PROJECT ASSESSMENT MANUAL Rut-Depth Measurements NCHRP Synthesis 334 (McGhee, 2004) notes that 46 state departments of transporta- tion (DOTs) collect automated rut-depth measurements almost always associated with roughness measurements. McGhee (2004) and SHRP (1993) define rut depth as the “longitudinal surface depressions in the wheelpaths.” Figure 1.10 helps to define lateral locations of a typical highway lane (from AASHTO, 2001). Figure 1.11 shows how rut depths are measured with automated equipment. Figure 1.10. Wheelpaths and area between wheelpaths. Source: McGhee, 2004, and AASHTO, 2001. Figure 1.11. Rut-depth measurements. Source: McGhee, 2004, and AASHTO, 2000.

14 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE International Roughness Index (IRI) Measurements McGhee (2004) defines pavement roughness as the “deviation of a surface from a true planar surface with characteristic dimensions that affect vehicle dynamics and ride quality.” The Standard Practice for Computing International Roughness Index of Roads from Longitudinal Profile Measurements (ASTM E1926-08) defines the inter- national roughness index (IRI) as the “pavement roughness index computed from a longitudinal profile measurement using a quarter-car simulation at a simulation speed of 80 kph (50 mph).” Further, ASTM E1926 notes that “IRI is reported in either meters per kilometer (m/km) or inches per mile (in/mile).” Analysis Tools Some of the analysis tools available include allowable rut depths and recommended IRI levels, which are shown in Tables 1.11 through 1.13. A study done in Wisconsin found, for state highways with speed limits greater than 45 mph, hydroplaning-related accidents significantly increased when rut depths were 0.3 in. or greater (Start, Jeong, and Berg, 1998). State DOTs such as the Washington State DOT (WSDOT) use a rehabilitation trigger level of 0.4 in. (10 mm). The Texas DOT notes in its Hydraulic Design Manual (2009) that water depths of 0.2 in. or greater, and Fwa (2006) found that a rut depth of 0.5 in. or more, can create the poten- tial for hydroplaning. Thus, a rut depth greater than or equal to 0.5 in. appears to be a reasonable trigger level for rehabilitation decisions. TABLE 1.11. TYPICAL MAXIMUM RUT DEPTHS Pavement Type Maximum Rut Depth, in. (mm) Texas DOT [concern about hydroplaning] >0.2 (>5) Wisconsin Hydroplaning Study (Start et al., 1998) 0.3 (7.6) Washington State DOT 0.4 (10) Fwa (2006) [based on hydroplaning] 0.5 (12.5) Shahin (1997) [from the PAVER Asphalt Distress Manual—Pavement Distress Identification Guide for Asphalt-Surfaced Roads and Parking Lots] Low 0.25–0.5 (6–13) Medium 0.5–1.0 (13–25) High >1.0 (>25) The IRI criteria used by FHWA have evolved as illustrated by review of Tables 1.12 and 1.13. The most detailed breakdown was distributed by FHWA in 1999 and sug- gests that IRI values of less than 60 in./mile are quite good, whereas those greater than 170 in./mile are poor. Interestingly, many newly paved HMA projects typically have IRI values close to the 60 in./mile value. Eventually, FHWA simplified its criteria as shown in Table 1.12.

15 PROJECT ASSESSMENT MANUAL A study conducted on Seattle-area urban freeways using driver in-vehicle opinion surveys (Shafizadeh and Mannering, 2003) confirmed that motorists find pavements with IRI values less than 170 in./mile acceptable as to ride quality (85% acceptable). The paper concluded that there was no evidence to change federal IRI guides (in essence those shown in Table 1.13). TABLE 1.12. FHWA IRI CRITERIA Ride Quality Terms All Functional Classifications IRI, in./mi (m/km) PSR Rating Good <95 (<1.5) Good Acceptable ≤170 (≤2.7) Acceptable Not acceptable >170 (>2.7) Not acceptable Source: FHWA, 2006. TABLE 1.13. EARLIER FHWA IRI CRITERIA Ride Quality Terms PSR Rating IRI, in./mile (m/km) National Highway System Ride Quality Very good ³4.0 <60 (<0.95) Acceptable, between 0 and 170 in./mileGood 3.5–3.9 60–94 (0.95–1.48) Fair 3.1–3.4 95–119 (1.50–1.88) Mediocre 2.6–3.0 120–170 (1.89–2.68) Poor ≤2.5 >170 (>2.68) Less than acceptable, >170 in./mile Source: FHWA, 1999. NONDESTRUCTIVE TESTING VIA THE FALLING WEIGHT DEFLECTOMETER (FWD) Purpose This section overviews the most commonly used FWD in use and how it can be used to aid pavement assessment decisions. Measurement Method This subsection briefly overviews impact (or impulse) pavement loading. FWD devices can obtain measurements rapidly and the impact load is easily varied. All impact load nondestructive testing (NDT) devices deliver a transient impulse load to the pavement surface. The subsequent pavement response (deflection) is mea- sured. Standard test methods include the following: • ASTM D4694-96: Standard Test Method for Deflections with a Falling-Weight- Type Impulse Load Device. • ASTM D4695-03: Standard Guide for General Pavement Deflection Measurements.

16 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE The significant features of ASTM D4694 include the following: (1) the force pulse will approximate a haversine or half-sine wave; (2) the peak force of 11,000 lb must be achievable by the loading device; (3) the force-pulse duration should be within the range of 20 to 60 ms with a rise time in the range of 10 to 30 ms; (4) the loading plate standard sizes are 300 mm (11.8 in.) and 450 mm (18 in.); (5) the deflection transducers, which are used to measure the maximum vertical movement of the pavement, can be seis mometers, velocity transducers, or accelerometers; (6) the load measurements must be accurate to at least ±2% or ± 160 N (±36 lb), whichever is greater; (7) the deflection measurements must be accurate to at least ±2% or ±2 µm (±0.08 mil), whichever is greater (note that 0.08 mil = 0.00008 in. and 2 µm = 0.002 mm); and (8) a precision guide in ASTM D4694 notes that, when a device is operated by a single operator in repetitive tests at the same location, the test results are questionable if the difference in the measured center deflection (D 0 ) between two consecutive tests at the same drop height (or force level) is greater than 5%. For example, if D 0 = 0.254 mm (10 mils) then the next load must result in a D 0 range less than 0.241 mm to 0.267 mm (9.5 to 10.5 mils). Falling Weight Deflectometer (FWD) The FWD is widely used in the United States. The device is trailer-mounted and uses deflection sensors that are velocity transducers. By use of different drop weights and heights this device can vary the impulse load to the pavement structure from about 1,500 to 27,000 lb. The weights are dropped onto a rubber buffer system resulting in a load pulse of 0.025 to 0.030 s. The standard load plate has a 300-mm (11.8-in.) diameter. Locations for the seven to nine velocity transducers vary. According to ASTM D4694, the number and spacing of the sensors is optional and will depend upon the purpose of the test and the pavement layer characteristics. A sensor spacing of 12 in. is frequently used. A number of state DOTs have used 0, 8, 12, 24, 36, and 48 in. for the distance from the center of the load plate. SHRP used sensor spacings of 0, 8, 12, 18, 24, 36, and 60 in. from the center of an 11.8-in. load plate. Analysis Tools This subsection focuses on straightforward analysis tools that can be applied to FWD deflection results. Description of Available Analysis Tools for Flexible Pavements This subsection describes three data-assessment tools: (1) maximum deflection, (2) the area parameter, and (3) a simplified method for calculating the subgrade modulus. The use of selected indices and algorithms provides a “picture” of the relative con- ditions found throughout a project. This picture is useful in performing back calculation and may at times be used by itself on projects with large variations in surfacing layers. Deflections measured at the center of the test load combined with area values and ESG computed from deflections measured at 24 in. from the center of the load plate are shown in the linear plot to provide a visual picture of the conditions found along the length of any project (as illustrated by data from a rural road in Figure 1.12).

17 PROJECT ASSESSMENT MANUAL The deflection data in Figure 1.12 are “normalized” data in that the measured deflections are calculated for a 9,000-lb load. The modulus determination was based on the deflection 24 in. from the center of the load plate. Table 1.14 provides general information about conclusions that can be drawn from the FWD parameters of area and D0. TABLE 1.14. GENERAL INFORMATION ABOUT THE AREA AND D0 FWD-Based Parameter Generalized ConclusionsArea Maximum Surface Deflection (D0) Low Low Weak structure, strong subgrade Low High Weak structure, weak subgrade High Low Strong structure, strong subgrade High High Strong structure, weak subgrade Figure 1.12. Illustrations of FWD deflection data summarized by the three types of data.

18 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Maximum Pavement Deflection (D0) The maximum pavement deflection can vary widely for different pavement structures and throughout the day as the temperature changes. Ranges of D0 can be grouped into the broad and approximate categories shown in Table 1.15. TABLE 1.15. D0 RANGES Maximum Surface Deflection (D0) Level Generalized Conclusions Approximate D0 (in.) Low deflections Strong structure ≤0.020 Medium deflections Medium structure 0.030 High deflections Weak structure >0.050 Area Parameter The area parameter represents the normalized area of a slice taken through any deflec- tion basin between the center of the test load and 3 ft. ”Normalized” means that the area of the slice is divided by the deflection measured at the center of the test load, D 0 . Thus, the area parameter is the length of one side of a rectangle where the other side of the rectangle is D 0 ; hence, the area parameter has units of inches. The area equation is: A = 6(D0 + 2D1 + 2D2 + D3)/D0 where D0 = surface deflection (mils) at the center of the test load, D1 = surface deflection (mils) at 1 ft, D2 = surface deflection (mils) at 2 ft, and D3 = su.rface deflection (mils) at 3 ft. The maximum value for area is 36.0 and occurs when all four deflection measure- ments are equal (not likely to actually occur), as follows: If D 0 = D 1 = D 2 = D 3, then area = 6(1 + 2 + 2 + 1)/1 = 36.0 m = 36.0 in. All four deflection measurements being equal (or nearly equal) indicates an extremely stiff pavement system (like PCC slabs or thick, full-depth asphalt concrete). The minimum area value should be no less than 11.1 in. This value can be calcu- lated for a one-layer system, which is analogous to testing (or deflecting) the top of the subgrade (i.e., no pavement structure). Using appropriate equations, the ratios D D D D D D , ,1 0 2 0 3 0

19 PROJECT ASSESSMENT MANUAL always result in 0.26, 0.125, and 0.083, respectively. Putting these ratios in the area equation results in area = 6(1+ 2(0.26) + 2(0.125) + 0.083)/1 = 11.1 in. Further, this area value suggests that the elastic moduli of any pavement system would all be equal (e.g., E1 = E2 = E3 = …). This is highly unlikely for actual, in-service pavement struc- tures. Low area values suggest that the pavement structure is not much different from the underlying subgrade material (which is not always a bad thing if the subgrade is extremely stiff). Typical area values are shown in Table 1.16. TABLE 1.16. TYPICAL AREA VALUES Pavement Structure Area Parameter (in.) PCC pavement range 24–33 “Sound” PCC 29–32 Thick HMA (~9 in. of HMA) 27+ Medium HMA (~5 in. of HMA) 23 Thin HMA (~2 in. of HMA) 17 Chip-sealed flexible pavement 15–17 Weak chip-sealed flexible pavement 12–15 Subgrade Modulus An NCHRP study (Darter, Elliott, and Hall, 1991), which revised Part III of the AASHTO Pavement Guide, recommended that the following equation be used to solve for subgrade modulus: MR = P(1 − µ2)/(m)(Dr)(r) (1.1) where MR = backcalculated subgrade resilient modulus (psi), P = applied load (lb) from the FWD, Dr = pavement surface deflection a distance r from the center of the load plate (in.), and r = distance from center of load plate to Dr (in.). Using a Poisson ratio of 0.40, Equation 1.1 reduces to MR = 0.01114(P/D2), (1.2) MR = 0.00743(P/D3), (1.3) MR = 0.00557(P/D4), (1.4) for sensor spacing of 2 ft (610 mm), 3 ft (914 mm), and 4 ft (1219 mm), respectively.

20 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE If a Poisson ratio of 0.45 is used instead for the same sensor spacing, the equations become MR = 0.01058(P/D2), (1.5) MR = 0.00705(P/D3), (1.6) MR = 0.00529(P/D4). (1.7) Darter et al. (1991) recommended that the deflection used for subgrade modu- lus determination should be taken at a distance at least 0.7 times r/ae, where r is the radial distance to the deflection sensor and ae is the radial dimension of the applied stress bulb at the subgrade “surface.” The ae dimension can be determined from the following: a a D E Me P R 2 3 2 = +     where ae = radius of stress bulb at the subgrade–pavement interface, a = NDT load plate radius (in.), D = total thickness of the pavement layers (in.), EP = effective pavement modulus (psi), and MR = backcalculated subgrade resilient modulus. For “thin” pavements, ae 15 in., and for “medium” to “thick” pavements, ae 26 to 33 in. Thus, the minimum r is usually 24 to 36 in. (recall r > 0.7(ae)). Typical subgrade moduli are shown in the Table 1.17 (after Chou, Uzan, and Lytton, 1989). TABLE 1.17. TYPICAL SUBGRADE MODULI Material Subgrade Moduli and Climate Condition Dry (psi) Wet, No Freeze (psi) Wet, Freeze Unfrozen (psi) Frozen (psi) Clay 15,000 6,000 6,000 50,000 Silt 15,000 10,000 5,000 50,000 Silty or clayey sand 20,000 10,000 5,000 50,000 Sand 25,000 25,000 25,000 50,000 Silty or clayey gravel 40,000 30,000 20,000 50,000 Gravel 50,000 50,000 40,000 50,000

21 PROJECT ASSESSMENT MANUAL Example Analyses of FWD Deflection Basins for Flexible Pavement The deflection basins shown in Table 1.18 were obtained with an FWD. The pavement temperature at the time of testing was 46°F (8°C). The deflection basins for the four FWD drops normalized to 9,000 lb are shown in the table. TABLE 1.18. EXAMPLE FWD DEFLECTION DATA FWD Load (lb) Deflection (mils) D0 D8˝ D12˝ D24˝ D36˝ D48˝ 16,987 27.07 21.55 18.60 11.27 7.33 5.28 12,070 21.28 16.98 14.62 8.67 5.56 3.98 9,406 17.53 13.95 11.98 7.01 4.45 3.23 6,186 12.33 9.77 8.31 4.65 2.88 2.05 Normalized to 9,000 lb 16.59 13.24 11.34 6.58 4.18 2.99 The pavement structure at the time of FWD testing was as follows: • HMA: 6.0 in. (and the HMA layer exhibited some fatigue cracking). • Granular base (sandy gravel): 18.0 in. • Subgrade: Silt (ML) with a wide seasonal variation in water table depth. The soil is frost susceptible and this area can have substantial ground freezing. The spring thaw occurred about one month prior to the testing. Requirements Analyze the available data to characterize the overall structure and estimate the layer properties (moduli) using only the information provided above. Results Maximum surface deflection The maximum surface deflection is 0.01659 in. for a pavement with 6 in. of HMA. This value suggests a “low” pavement deflection. Subgrade modulus (closed-form equations) from the AASHTO Guide (1993) MR = P(1−µ2)/(p)(Dr)(r) = 9000(1−0.452)/(p)(0.00418)(36) ≅ 15,200 psi. Check r ≥ 0.7(ae), OK. The pavement subgrade modulus for an ML silt is better than average.

22 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Area parameter Area = 6(D0 + 2D12˝ + 2D24˝ + D36˝)/D0 = 6(0.01659 + (2)(0.01134) + (2)(0.00658) + 0.00418)/0.01659 ≅ 20.5 in. This area parameter is low for this thickness of asphalt concrete (AC). Thus, the area value suggests a weak pavement structure but not extremely so. More Detailed Project Data Example Table 1.19 summarizes deflection data that were collected on a portion of an actual project. The project was about 5 mi in length and FWD testing was performed every 250 ft, but only four of the FWD locations are shown (these locations were also cor- ing sites). The average pavement temperature at the time the FWD data were collected was 46°F to 50°F. The timing of the survey was about 1.5 to 2 months after the spring thaw in this area. TABLE 1.19. FWD DEFLECTIONS, AREA VALUE, AND SUBGRADE MODULUS Core Location (MP) Load (lbf) Deflections (mils) Area Value (in.) MR (psi)D0 D8˝ D12˝ D24˝ D36˝ D48˝ 207.85 16,940 31.30 26.18 23.19 13.78 9.09 6.65 12,086 24.21 20.31 18.11 10.35 6.81 4.96 9,421 19.45 16.38 14.57 8.11 5.28 3.98 6,218 13.19 11.26 9.92 5.12 3.39 2.83 Normalized Values 18.39 15.51 13.78 7.60 5.00 3.82 21 14,358 208.00 16,987 27.04 21.53 18.58 11.26 7.32 5.28 12,070 21.26 16.97 14.61 8.66 5.55 3.98 9,405 17.52 13.94 11.97 7.01 4.45 3.23 6,186 12.32 9.76 8.31 4.65 2.87 2.05 Normalized Values 16.57 13.23 11.34 6.57 4.17 2.99 20 16,534 208.50 16,829 14.92 11.89 10.23 5.91 3.19 2.28 12,245 11.65 9.29 7.95 4.49 2.13 1.73 9,533 9.61 7.63 6.53 3.62 1.81 1.30 6,297 6.73 5.35 4.49 2.40 1.26 0.87 Normalized Values 9.01 7.17 6.10 3.39 1.69 1.26 19 32,198 209.00 16,305 59.25 48.58 42.52 21.30 9.53 5.12 11,737 46.14 37.52 32.56 15.59 6.69 3.58 9,247 36.93 29.80 25.63 11.77 4.96 2.68 6,154 25.00 19.88 16.77 7.28 3.03 1.73 Normalized Values 35.51 28.66 24.61 11.42 4.84 2.64 19 9,572

23 PROJECT ASSESSMENT MANUAL As shown in Table 1.19, the normalized D0 deflections range from about 9 to 36 mils. Deflections less than about 30 mils are considered normal. The HMA thick- nesses varied between 4.6 and 5.3 in. with an average of 5.2 in., which constitutes a “medium” thickness of HMA (refer back to Table 1.15). The area values shown in the table suggest weak HMA, but not necessarily extreme weakness due to stripping. Table 1.20 illustrates typical theoretical area values for various uncracked HMA thicknesses, which aids this type of comparison. TABLE 1.20. TYPICAL THEORETICAL AREA VALUES FOR UNCRACKED HMA HMA Thickness (in.) Approximate Area Parameter (in.) Normal Stiffness Low Stiffness 2 17 16 3 19 18 4 21 19 5 23 21 6 24 22 7 26 22 8 26 23 9 27 24 10 28 24 A quick, slightly more formal check of the pavement structure is to compare the actual area value to see if it falls within the range (normal to low stiffness), above the range (above normal stiffness), or below the range (below normal stiffness). This com- parison is shown in Table 1.21. TABLE 1.21. COMPARISON OF AREA VALUE AND ACCEPTABLE AREA VALUE RANGE Core Location HMA Thickness (in.) Actual Area (in.) Above, Below, or Within Range 207.85 5.3 21 Within 208.00 6.0 20 Below 208.50 4.7 19 Below 209.00 4.6 19 Below Area values provide a very good check on whether the surface cracking observed is full-depth or top-down cracking. If it is top-down cracking, then the area values will be within the expected range. If it is full-depth cracking, then the area values will be well below the expected range. The area values may also be used to provide some assess- ment of changes in pavement structure or depth, and as such the measure provides a good basis for coring locations. And finally, the area values can provide a very good check against the relative values for the backcalculated modulus for the HMA layer.

24 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Description of Available Analysis Tools for Rigid Pavements Rehabilitation of PCC pavements is not straightforward. To provide a more consis- tent analysis process, the load transfer efficiency (LTE) should be checked with FWD- obtained deflection data if the pavement type is JPCP. Load Transfer Efficiency When a wheel load is applied at a joint or crack, both the loaded slab and the adjacent unloaded slab deflect. The amount the unloaded slab deflects is directly related to joint performance. If a joint is performing perfectly, the loaded and unloaded slabs deflect equally. Joint performance can be evaluated by calculating LTE across a joint or crack using measured deflection data. The concept of joint load transfer efficiency is illus- trated in Figure 1.13. LTE can be calculated using the following equation: LTE = (∆U/DL)(100) where LTE = load transfer efficiency (%), ∆U = deflection of the unloaded slab (mils), and DL = loaded slab deflection (mils). Figure 1.13. Illustration of joint load transfer efficiency. Source: NHI, 2003.

25 PROJECT ASSESSMENT MANUAL Figure 1.14. Locations of FWD load plate and deflection sensors for determining load transfer efficiency. Source: NHI, 2003. Joint efficiency depends on several factors, including temperature (which affects joint opening), joint spacing, number and magnitude of load applications, foundation sup- port, aggregate particle angularity, and the presence of mechanical load transfer devices. As mentioned, temperature plays a major role in determining joint effectiveness. In general, the lower the temperature, the lower the load transfer efficiency. Load trans- fer efficiency is reduced because joints open during cooler weather, reducing contact between faces. Joint load transfer efficiency has also been shown, in both laboratory and field studies, to decrease with increasing load applications. However, this impact is lessened for harder aggregates. The aggregate characteristics play a more significant role after many load applications. To test the approach side of a joint or crack, the FWD loading plate is placed in front of the joint, with the other velocity transducers located across the joint. The leave side of the joint is tested by placing the loading plate at the joint edge on the leave slab with an extra velocity transducer mounted behind the loading plate across the joint. The concept of slab approach and leave sides and of transverse joint testing are illus- trated in Figure 1.14.

26 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE The percentage of load transfer can vary between almost 100% (excellent) to near 0% (extremely low). AASHTO (1993) notes that load transfer restoration should be considered for transverse joints and cracks with load transfer efficiencies ranging from 0% to 50%. It has been observed for numerous in-service jointed PCC pavements that load transfer efficiencies of 70% or greater generally provide good joint or crack performance. Backcalculation Backcalculation is the process by which pavement layer moduli are estimated by match- ing measured and calculated surface deflection basins. This is done via a computer program and there are a number of these available in the United States. It is likely that within a specific state there is a preferred backcalculation software package to use. The general guidelines that follow are broad in scope and should be considered “rules of thumb.” Number of Layers Generally, use no more than three or four layers of unknown moduli in the back- calculation process (preferably, no more than three layers). If a three-layer system is being evaluated, and questionable results are being produced (e.g., extremely weak base moduli), then it is sometimes advantageous to evaluate this pavement structure as a two-layer system. Some experienced users have found that there are times, such as dealing with a highly stress-sensitive subgrade, when it may be beneficial to consider adding layers to reduce compensating error effects. This modification would possibly indicate the base material has been contaminated by the underlying subgrade and is weaker due to the presence of fine material. Alternatively, a stiff layer should be con- sidered if it has not been considered previously (see below). If a pavement structure consists of a stiffer layer between two weak layers, it may be difficult to obtain realistic backcalculated moduli. An example is a pavement structure that consists of deterio- rated asphalt concrete over a cement-treated base. A stiff underlying layer, if found, is typically given a modulus value and is treated as a layer of known moduli. Thickness of Layers Surfacing. It can be difficult to “accurately” backcalculate HMA or BST moduli for bituminous surface layers less than 3 in. thick. Such backcalculation can be attempted for layers less than 3 in. thick, but caution is suggested. In theory, it is possible to backcalculate separate layer moduli for various types of bituminous layers within a flexible pavement. Generally, it is not advisable to do this because one can quickly attempt to backcalculate too many unknown layer moduli (i.e., more than three or four). By necessity, one should expect to combine all bitumi- nous layers (seal coats, asphalt concrete, etc.) into “one” layer unless there is evidence (or the potential) of distress, such as stripping, in an HMA layer, or some other such distress that is critical to pavement performance.

27 PROJECT ASSESSMENT MANUAL Unstabilized base and subbase course. A “thin” base course beneath “thick” surfacing layers (say, HMA or PCC) often results in low base moduli. There are a number of reasons why this can occur. First, a thin base is not a “significant” layer under a stiff, thick layer and, where the measured surface deflections are not significantly affected by the layer, its moduli cannot be backcalculated. Second, the base modulus may be relatively “low” due to the stress sensitivity of granular materials. The use of a stiff layer generally improves the modulus estimate for base and subbase layers. Subgrade If unusually high subgrade moduli are calculated, a check should be performed to see if a stiff layer is present. Stiff layers, if unaccounted for in the backcalculation process, will generally result in unrealistically high subgrade moduli accompanied by inappro- priately low base-layer moduli due to compensating error effects. This is particularly true if a stiff layer is within a depth of about 20 to 30 ft below the pavement surface. Stiff Layer Often, stiff layers are given “fixed” stiffness ranging from 100,000 to 1,000,000 psi with semi-infinite depth. This, in effect, gives the “subgrade” a layer with a “fixed” depth instead of the normally assumed semi-infinite depth. What is not so clear is whether one should always fix the depth to the stiff layer at 20, 30, or 50 ft if no stiff layer is otherwise indicated (i.e., use a semi-infinite depth for the subgrade). The depth to the stiff layer should be verified whenever possible with other NDT data or bor- ings. It should be noted that a number of backcalculation programs include a tool to estimate the depth to the stiff layer. The stiffness (modulus) of the stiff layer can vary. If the stiff layer occurs because of saturated conditions (e.g., water table), then moduli of about 50,000 psi appear more appropriate. If rock or stiff glacial tills are the source of the stiff layer, then moduli of about 1,000,000 psi appear to be more appropriate. Layer Moduli A few comments about layer moduli are appropriate. Cracked HMA moduli. Generally, fatigue-cracked HMA (about 10% wheel- path cracking) is often observed to have backcalculated moduli of about 100,000 to 250,000 psi. What is most important in the backcalculation process, assuming surface fatigue cracking is present, is to determine whether the cracks are confined to only the immediate wearing course or if they penetrate through the whole depth of the HMA layer. For HMA layers greater than 6 in. thick, cracking only in the wearing course is often observed and the overall HMA layer will have a substantially higher stiffness than noted above (at moderate layer temperatures of 75°F to 80°F). Base and subbase moduli. Typical base and subbase moduli are shown in Table 1.22.

28 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Subgrade moduli. Typical subgrade moduli were previously shown in Table 1.18. Backcalculation Summary Performing backcalculation of pavement layer moduli is part science and part art; thus, experience typically will improve the estimated results. It is advisable to initially work with someone who has solid experience doing backcalculation or to take a short course on the topic, assuming one is available. It will take only a few projects, along with experience from others, to become well informed about this powerful assessment technique. Reviews of Stubstad et al. (2006), Von Quintus and Simpson (2002), and Yau and Von Quintus (2002) provide additional insight into layer moduli and how to estimate them. These references cover work associated with the LTPP experiments and include both backcalculation and closed-form equations for developing moduli estimates along with laboratory results. GROUND-PENETRATING RADAR (GPR) Purpose This section describes GPR technology and presents an overview of the most common applications of both air-coupled and ground-coupled GPR systems for aiding in pave- ment assessment decisions. TABLE 1.22. TYPICAL UNSTABILIZED AND STABILIZED BASE AND SUBBASE MODULI Material Compressive Strength (psi) Typical Modulus (psi) Modulus Range (psi) Unstabilized Crushed stone or gravel base — 35,000 10,000–150,000 Crushed stone or gravel subbase — 30,000 10,000–100,000 Sand base — 20,000 5,000–80,000 Sand subbase — 15,000 5.000–80,000 Stabilized Lime stabilized <250 30,000 5,000–100,000 250–500 50,000 15,000–150,000 >500 70,000 20,000–200,000 Cement stabilized <750 400,000 100,000–1,500,000 750–1,250 1,000,000 200,000–3,000,000 >1,250 1,500,000 300,000–4,000,000

29 PROJECT ASSESSMENT MANUAL Measurement Method This section briefly describes the two types of GPR and the basic principles of operation. The standard references for GPR applications in highways are: • AASHTO PP 40-00: Standard Recommended Practice for Application of Ground Penetrating Radar to Highways. • ASTM D6087-08: Standard Test Method for Evaluating Asphalt Covered Con- crete Bridge Decks using Ground Penetrating Radar. • ASTM D6432-99 (2005): Standard Guide for Using Surface Ground Penetrating Radar Method for Subsurface Investigation. Air-Coupled GPR Systems A typical commercially available 2.2-GHz air-coupled GPR unit is shown in Figure 1.15. The radar antenna is attached to a fiberglass boom and suspended about 5 ft from the vehicle and 14 in. above the pavement. This particular GPR unit can op- erate at highway speeds (70 mph); it transmits and receives 50 pulses per second and can effectively penetrate to a depth of around 20 to 24 in. All GPR systems include a distance- measuring system and many of the new systems also have synchronized or integrated video logging, so the operator can view both surface and subsurface condi- tions. Global positioning is also included in many new systems for identifying problem locations. Figure 1.15. Air-coupled GPR systems for highways. Photo: Tom Scullion.

30 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE The advantages of these systems include the speed of data collection, which does not require any special traffic control. The GPR generates clean signals that without filtering are ideal for quantitative analysis using automated data-processing techniques to compute layer dielectrics and thicknesses. These systems are also excellent for locat- ing near-surface defects in flexible pavements. The disadvantages are that (a) they have a limited depth of penetration, (b) they are not ideal for penetrating thick concrete pavements, and (c) the most popular operat- ing frequency (1 GHz) is now subject to Federal Communications Commission (FCC) restrictions in the United States. Ground-Coupled GPR systems As shown in Figure 1.16, a whole range of different operating frequencies is available for ground-coupled GPR systems. The selection of the best frequency for a particular application depends on the required depth of penetration. As the name implies, these antennas have to stay in close contact with the pavement under test. The advantage of these systems is their depth of penetration; several of the lower- frequency systems can penetrate 20 ft under ideal conditions. The higher-frequency systems are superior for many concrete pavement applications such as locating both reinforcing steel and subslab defects such as voids or trapped moisture. The disadvan- tage of these systems is the speed of data collection; when towed behind a vehicle, the maximum speed is around 5 mph. The signals are also noisy, and filtering is required. Substantial training is required to clean up and interpret ground-coupled GPR data. Figure 1.16. Ground-coupled systems: 1.5 GHz (left); lower-frequency antennas with control unit (right). Photos: Tom Scullion.

31 PROJECT ASSESSMENT MANUAL Analysis Tools All GPR systems send discrete pulses of radar energy into the pavement and capture the reflections from each layer interface within the structure. Radar is an electromag- netic (EM) wave and therefore obeys the laws governing reflection and transmission of EM waves in layered media. At each interface within a pavement structure, a part of the incident energy will be reflected and a part will be transmitted. It is normal to collect between 30 and 50 GPR return signals per second, which for high-speed air- coupled surveys could mean a trace for every 2 to 3 ft of travel. The captured return signal is often color coded and stacked side-by-side to provide a profile of subsurface conditions; this is analogous to an “x-ray” of the pavement structure. Examples of this process are given later. However, with air-coupled signals as described below, these signals can also be used to automatically calculate the engineering properties of the pavement layers. Air-Coupled GPR System A typical plot of captured reflected energy versus time for one pulse of an air-coupled GPR system is shown in Figure 1.17 as a graph of volts versus arrival time in nano- seconds. To understand GPR signals, it is important to understand the significance of this plot. Figure 1.17. Captured GPR reflections from a typical flexible pavement.

32 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE The reflection, A0, is known as the end reflection and is internally generated system noise which is present in all captured GPR waves. The more important peaks are those that occur after A0. The reflection A1 (in volts) is the energy reflected from the surface of the pavement and A2 and A3 are reflections from the top of the base and subgrade, respectively. These are all classified as positive reflections, which indicates an interface with a transition from a low to a high dielectric material (typically low to higher mois- ture content). These amplitudes of reflection and the time delays between reflections are used to calculate both layer dielectrics and thickness. The dielectric constant of a material is an electrical property that is most influenced by moisture content and den- sity; it also governs the speed at which the GPR wave travels in the layer. An increase in moisture will cause an increase in layer dielectric; in contrast, an increase in air void content will cause a decrease in layer dielectric. The equations to calculate surface-layer thickness and dielectrics are summarized below: A A A A 1 / 1 /a m m 1 1 2 ∈ = + −     (1.8) where ∈a = dielectric of the surface layer, A1 = amplitude of surface reflection, in volts (V), and Am = amplitude of reflection from a large metal plate (V) (this represents the 100% reflection case; see Figure 1.15 for the metal plate test). h cx t a 1 1 = ∆ ∈ (1.9) where h1 = thickness of the top layer, c = constant speed of EM wave in air (5.9 in./ns two-way travel), and ∆t1 = time delay between peaks A1 and A2 (ns). Similar equations are available for calculating the base-layer dielectric and thick- ness. This calculation process is performed automatically in most operating systems with the end user simply obtaining a table of layer properties. In most GPR projects, several thousand GPR traces like Figure 1.17 are collected. To conveniently display and interpret this information, color-coding schemes are used to convert the traces into line scans and they are stacked side-by-side so that a sub- surface image of the pavement structure can be obtained. This approach is shown in Figure 1.18.

33 PROJECT ASSESSMENT MANUAL The raw GPR image collection is displayed vertically in the middle of Figure 1.18. This image is for one specific location in the pavement. The GPR antenna shoots straight down and the resulting thickness and dielectric estimates are point specific. The single trace generated is color coded into a line scan using the color scheme in the middle of Figure 1.18. In the current scheme, the high positive reflections are colored red and the negatives are colored blue. The green color is used where the reflections are near zero and are of little significance. These individual line scans are stacked so that a display for a length of pavement is developed. Being able to read and interpret these images is critical to effectively using GPR for pavement investigations, to locate section breaks in the pavement structure, and to pinpoint the location of subsurface defects. An example of a typical GPR display for approximately 3,000 ft by 24 in. deep of a thick flexible pavement is shown in Figure 1.19. This is taken from a section of newly constructed thick asphalt pavement over a thin granular base. In all such displays the x axis is distance (in miles and feet) along the section and the y axis is a depth scale in inches. Figure 1.18. Color coding and stacking individual GPR images. Source: Tom Scullion.

34 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE The labels on Figure 1.19 are as follows: (A) GPR files being used in analysis, (B) main pull-down menu bar, (C) button to define the color-coding scheme, (D) dis- tance scale (miles and feet), (E) end location of data within the GPR file (1 mi and 3,479 ft), (G) depth scale in inches, with the zero (0) being the surface of the pavement, and (F) default dielectric value used to convert the measured time scale into a depth scale. The important features of this figure are the lines marked H, I, and J; these are the reflections from the surface, top, and bottom of the base, respectively. This pave- ment is homogeneous and the layer interfaces are easy to detect. When processing GPR data, the first step is to develop displays such as Figure 1.19. From this, it is possible to identify any clear breaks in pavement structure and to iden- tify any significant subsurface defects. The intensity of the subsurface colors is related to the amplitude of reflection; therefore, areas of wet base would be observed as bright red reflections (I). For many applications, a black-and-white coding scheme is selected. This is widely used during review of data collected with ground-coupled GPR systems. An example of the grayscale image for the pavement shown in Figure 1.19 is shown in Figure 1.20. Figure 1.19. Color-coded GPR traces.

35 PROJECT ASSESSMENT MANUAL Figure 1.20. Data similar to Figure 1.19 presented as a grayscale image. All the commercially available software packages produce both a color display of the subsurface condition, such as Figures 1.19 and 1.20, together with a table of computed layer thicknesses and dielectrics that is usually exported to Excel. A typical table is shown in Figure 1.21, where Thick1 and E1 are the top-layer thickness and dielectric, respectively. Figure 1.21. Tabulated thicknesses and dielectric values from GPR data.

36 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Examples of Analysis of GPR Data for Flexible Pavements When planning to incorporate the existing pavement as part of a new pavement struc- ture, it is critical to have good information on the existing subsurface layer thicknesses and layer types. A few DOTs maintain good pavement layer databases, but this is not always the case. DOTs often have limited or inaccurate information on existing layer thicknesses. Often, maintenance activities significantly alter the as-constructed pave- ment structure in localized areas and these activities are often not captured in existing databases. One popular method of rehabilitating flexible pavements is by the use of full-depth reclamation (FDR) and chemical treatment to incorporate and stabilize the existing pavement to form a solid foundation layer for the new pavement structure. How- ever, because of the failure to account for the variability of the existing pavement in the design phase, several major problems have occurred during construction, or poor pavement performance has resulted. Laboratory designs are based on testing at local- ized sampling locations, which can miss discrete areas of variable thickness. GPR can help address this issue. It also must be recalled that processing FWD data, as described in the section “Nondestructive Testing via the Falling Weight Deflectometer (FWD),” requires infor- mation about the thickness of the asphalt surface layer. GPR can provide substantial help in analyzing and explaining FWD deflection data. Three case studies are presented below to demonstrate how GPR can assist in flex- ible pavement evaluations. Thickness Profiling for an FDR Application In many FDR applications, the purpose is to treat the existing pavement to create a stable uniform pavement foundation layer for the new pavement structure. In most FDR applications, design samples are taken from the existing pavement and tested in the laboratory to determine the optimal level of either cement or asphalt stabilization to reach a specified target strength. It is therefore important to know that the sampling location selected is representative of the overall project. It is also important to assess if the selected design will be appropriate when variations in layer thicknesses occur. Figure 1.22 shows variations in asphalt layer thickness for an FDR candidate. At the sample location the structure was 5 in. of asphalt and 10 in. of granular base. Based on laboratory test results, the plan was to recycle to a depth of 10 in., blend- ing 50% asphalt and 50% existing base with 3% cement. However, from a review of Figure 1.22, the average 5 in. of HMA has several noticeable exceptions. The first 800 ft only has 3 in. of asphalt, which is not thought to be a concern. However, for about 2,000 ft, the total HMA thickness is over 12 in. From previous experience, the 3% cement treatment does not work with 100% reclaimed asphalt pavement (RAP). In these locations it was necessary to modify the construction plan, wherein 5 in. of the existing HMA was milled and replaced with 5 in. of new base. In that way, the FDR process can continue and in all locations the as-designed 50/50 blend can be treated with cement.

37 PROJECT ASSESSMENT MANUAL Figure 1.22. Surface thickness variations from GPR profiling on FM 550. Defect Detection Prior to Pavement Rehabilitation In many cases, long life of the existing flexible pavement can be achieved by simply adding a structural overlay to the existing structure. This process works well provided there are no major defects in the existing HMA layer or flexible base layer. GPR has shown that it can be used to detect stripping problems in HMA layers and areas where the existing base layer is holding moisture. It must be recalled that GPR traces are col- lected frequently at 2- to 3-ft intervals, so very precise location of defects is possible. The GPR color-coded profile shown in Figure 1.19 is from a thick HMA section with no defects. This should be contrasted with the GPR profile shown in Figure 1.23. This again is a thick HMA section, but in this case there are strong reflections from within Figure 1.23. Using GPR to identify defects in surface and base layers.

38 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE the HMA and very strong reflections from the bottom of the layer. The red and blue re- flections from within the HMA are associated with deteriorated areas where moisture is trapped. When these deteriorated areas are close to the surface, they can severely affect long-term performance. Although the presence of defects in either HMA or base layers can be easily detected by GPR, the severity of these defects will need to be confirmed by localized coring. This is valuable input to the pavement designer, who has to make a decision that affects the future anticipated performance of the proposed section. If the defects are very localized, then full-depth milling can be used in these areas. Section Uniformity With many older pavements, particularly those involving some form of pavement wid- ening, the existing pavement structure can be very variable. It is important to identify the different structures in order to explain the cause of current conditions and to de- sign future repairs. Such a case is shown in Figure 1.24. This is a 1.8-mi section and the entire sec- tion had received a thin overlay. However, the first part of the section was performing poorly. A GPR survey was undertaken and from the display it was clear that this sec- tion had three distinct pavement structures. Structure A was a thin HMA pavement over a flexible base, structure B was thick HMA, and structure C was a road built on top of an existing roadway. This type of subsurface mapping can clearly help designers with their rehabilitation designs. Figure 1.24. Using GPR to map subsurface variability.

39 PROJECT ASSESSMENT MANUAL Examples of Analysis of GPR Data for PCC Pavements The most popular applications of GPR in evaluating concrete pavements when making pavement rehabilitation decisions are (a) measuring slab thickness, (b) detecting the presence and depth of reinforcing steel, and (c) identifying problems beneath the slab such as voids or trapped moisture. In several instances, especially for steel detection, the ground-coupled systems perform better than the air-coupled systems. The high- frequency ground-coupled systems, such as the 1.5-GHz unit shown in Figure 1.16, can give more focus and better target resolution than air-coupled units. Several case studies are shown below. Rebar Detection The Geophysical Survey Systems, Inc. (GSSI) “Handbook for Radar Inspection of Concrete” (GSSI, 2006) has some good examples on rebar detection. Figure 1.25 shows the typical GPR signature obtained over reinforcing steel. There is a hyperbola shape and the top of the hyperbola is the location of the steel. The surface of the con- crete is the “direct couple” signature, and the depth between the surface and the top of the hyperbola is the depth of the concrete cover. GSSI also claims that the size of the rebar can be determined by the shape of the hyperbola. Figure 1.25. Ground-coupled GPR signals from steel in concrete. Source: GSSI. By moving the GPR antenna slowly across the surface of the concrete, it is pos- sible to map different layers of steel and the bottom of the concrete slab as shown in Figure 1.26.

40 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Void Detection Detecting thin air voids (which can be very detrimental to slab performance) with air- coupled GPR is often problematic. Controlled studies have found that air voids less than 0.75 in. thick cannot be readily detected with air-coupled GPR. However, if the voids are larger or if they are moisture filled, then they can readily be detected. An example of a GPR color profile for an 8-in. PCC slab with water-filled voids is shown in Figure 1.27. The strong reflections (red areas) indicate locations of trapped water. Figure 1.27. Mapping subslab water-filled voids with GPR. Figure 1.26. Mapping multiple layers of steel in concrete. Source: GSSI. Deep Investigations of Subslab Conditions with GPR Lower-frequency ground-coupled GPR can be used to investigate deep beneath con- crete pavements to identify changes in support conditions and possibly to help explain the occurrence of surface distress. Figure 1.28 shows the color profile from a 400-MHz

41 PROJECT ASSESSMENT MANUAL Figure 1.28. Mapping concrete pavement structure with GPR. ground-coupled system. The entire pavement system and changes in pavement support can be observed. The transverse rebar can be seen toward the top of the figure. The steel is more closely spaced in the left of the figure. The anomaly on the left is a culvert. The bottom of the slab is indicated. There is a clear change in subgrade support at the top of the subgrade, showing the transition from a cut to a fill area. Implementing GPR Technology for Pavement Evaluation GPR is an excellent technology for inspecting pavements when pavement rehabilita- tion decisions are being made. Many case studies have been presented over the past two decades, but widespread implementation of the technology has been slow. There are several factors causing this and they are discussed in this section. The main factors are the following: 1. The FCC banned 1-GHz air-coupled systems in 2002 (these units can be purchased in any country worldwide except the United States). For the past decade, most air-coupled GPR systems have been performed with systems built before 2002. Only recently have commercial systems such as GSSI’s 2.2-GHz system become available. 2. There is a lack of understanding about what GPR can and cannot do; in many cases the technology was oversold.

42 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE 3. Current data-processing software are inadequate, and there is a lack of end-user training. Agencies undertaking GPR implementation should be aware of the following issues, which must be resolved before GPR can be implemented as a routine pavement inspection tool. 1. Need for GPR hardware specifications, 2. Need for data-collection software specifications, 3. Training and specifications for data-collection activities, 4. Specifications and software for processing and interpreting GPR signals, 5. End-user training, and 6. Specifications for output formats and data storage systems. Several DOTs have implemented GPR technology in house (e.g., the Florida DOT and the Texas DOT), but most agencies get GPR services from consultant companies. Selecting the best vendor can also be a problem. Obtaining GPR Services The AASHTO publication has a short section with recommendations for agencies on hiring GPR consultants. In initiating contracts, the agency has to be convinced that • The consultant has quality equipment. The agency should ask the consultant to run their equipment against the performance specs (which are available). • The consultant has good data-processing skills. References from existing cus- tomers will help here. GPR interpretation should never be done without taking limited field verification cores early in the project. If the project is for layer-thick- ness determination or for defect detection, it should be simple to set up a verifica- tion system early in the project. Barriers to GPR Implementation In addition to the FCC requirements, there are also several common misconceptions that must be overcome before any agency will adopt GPR technology. These include the following: • GPR is only for layer-thickness determination: My state has good as-built records so we do not need GPR. As noted throughout this report, GPR is much more than a thickness-measuring tool. It provides information on the quality of existing structures and helps explain the causes of pavement distresses. Distresses are often associated with moisture ingress into pavement layers. GPR signals are highly sen- sitive to moisture in any layer. • GPR systems are too expensive. A complete air-coupled system described in this section costs around $100,000 for the complete turnkey system, including the vehicle. Ground-coupled systems cost approximately $60,000. Compared to the costs of pavement rehabilitation activities, GPR costs are minimal.

43 PROJECT ASSESSMENT MANUAL • GPR is a black box that is impossible to understand. This is not true; the basics of GPR are simple. The key here is that agency personnel should attend training schools to understand this technology. Even if the plan is to initiate GPR work through consultants, the agency personnel need to have a basic understanding of what this technology can and cannot do. • Our first experience with GPR was disappointing. This is often true. In the early 1990s, a host of companies sold GPR services. They sometimes made extensive claims on GPR’s potential and their ability to successfully interpret the signals. Many claimed to be able to find thin voids beneath concrete pavements, often to disappoint the DOT when validation field cores were taken. In some cases, the vendors did not have adequate software or interpretation skills. The key here again is training for end-user agency personnel. The AASHTO publication also is a good source to identify applications that have a high probability of success. • When the agency initiates a GPR program, a host of vendors make claims about their capabilities and it is impossible for the agency to judge their merits. This is often true, but it can be overcome by training of end-user agency personnel prior to initiating a program. Also, as with any new technology, field verification of any predictions must be a critical part of any program. GPR will not eliminate coring, but it will greatly reduce the number of cores. PAVEMENT CORES Purpose This section overviews pavement cores and how they can be used to aid pavement assess ment decisions. Much of pavement analysis and understanding stems from knowledge of layer thicknesses, types of materials, and condition. Measurement Method This subsection briefly overviews both the frequency of sampling and the organiza- tion of data from pavement cores. Pavement coring not only reveals much about the existing pavement structure, but it also allows for use of the DCP. Knowing the HMA layer thickness to within ¼ in. is essential in ensuring a more accurate predic- tion of layer moduli if a backcalculation procedure is used. The number of cores obtained will depend on project-specific conditions; how- ever, a reasonable rule of thumb is to obtain a core at every 5th or 10th FWD test location. If the pavement thicknesses are found to vary substantially (not probable, but this can be the case), then cores should be obtained at every FWD test location in those vicinities. FWD area values plotted along the project limits [as discussed in the section “Non destructive Testing via the Falling Weight Deflectometer (FWD)”] pro- vide good guidance for determining core locations because substantial changes in the pavement structure can be identified. If GPR data are collected, using the layer profiles in conjunction with FWD area values also provides very good guidance for develop- ing coring plans. Calibration cores for GPR data collection can also be used for other assessments.

44 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Typical core diameters are either 4 or 6 in. Coring should also be used to verify the depth of cracking (i.e., determination of top-down versus bottom-up cracking) as well as the presence and severity of stripping in HMA mixtures. Analysis Tools This subsection focuses on how to organize pavement core data to aid decision making. Core data should be organized similarly to the example data shown in Table 1.23. Additionally, the location of each core in the lane should be recorded (such as center- line, left wheelpath, between wheelpath, right wheelpath, and outside pavement edge). TABLE 1.23. ORGANIZATION OF PAVEMENT CORE DATA Core Location (milepost) Depth Comments (Cores should be taken frequently at cracks, if they exist, to determine if the crack is full depth or partial depth) HMA (in.) Base (in.) 207.85 5.3 18.0 Core taken at a crack, crack is full depth 208.00 6.0 18.0 Core taken at a crack, core not intact 208.50 4.7 12.0 Core taken at a crack, crack is full depth 209.00 4.6 12.0 Very fatigued, core broke into several pieces DYNAMIC CONE PENETROMETER (DCP) Purpose This section overviews the DCP and how it can be used to aid pavement assessment decisions. Measurement Method This subsection describes the DCP device. The standard test method is: • ASTM D6951-03: Standard Test Method for Use of the Dynamic Cone Penetrometer in Shallow Pavement Applications. From ASTM D6951: “This test method is used to assess in situ strength of undis- turbed soil and/or compacted materials. The penetration rate of the 8-kg DCP can be used to estimate in situ CBR (California Bearing Ratio), to identify strata thickness, shear strength of strata, and other material characteristics. The 8-kg DCP is held verti- cally and therefore is typically used in horizontal construction applications, such as pavements and floor slabs. This instrument is typically used to assess material proper- ties down to a depth of 1000-mm (39-in.) below the surface. The penetration depth can be increased using drive rod extensions. However, if drive rod extensions are used, care should be taken when using correlations to estimate other parameters since these cor- relations are only appropriate for specific DCP configurations. The mass and inertia of the device will change and skin friction along drive rod extensions will occur.”

45 PROJECT ASSESSMENT MANUAL “The 8-kg DCP can be used to estimate the strength characteristics of fine- and coarse-grained soils, granular construction materials and weak stabilized or modified materials. The 8-kg DCP cannot be used in highly stabilized or cemented materials or for granular materials containing a large percentage of aggregates greater than 50-mm (2-in.). The 8-kg DCP can be used to estimate the strength of in situ materials underly- ing a bound or highly stabilized layer by first drilling or coring an access hole.” An illustration of a standard DCP is shown in Figure 1.29. Figure 1.29. Minnesota DOT DCP. Source: Minnesota DOT, 1993.

46 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Analysis Tools DCP test results are typically expressed in terms of the DCP penetration index (DPI), which is the vertical movement of the DCP cone produced by one drop of the hammer. This is expressed as either mm/hammer blow or in./hammer blow (Minnesota DOT, 1993). Basic Correlation A common correlation with DCP data is to estimate the California bearing ratio (CBR) of unstabilized materials in a pavement structure. The following is a correlation devel- oped by the U.S. Army Corps of Engineers (Webster, Grau, and Thomas, 1992): log CBR = 2.46 – 1.12 log(DPI) or CBR = 292/DPI1.12 where DPI = mm/blow. Table 1.24 shows typical CBR and DPI ranges for three soil types (Minnesota DOT, 1993). TABLE 1.24. SOILS TYPES, CBR VALUES, AND DPI Soil Type CBR Range (%) DPI Range (mm/blow) Clay (CL) ∼1–14 15–127 Sand (S-W) 14–39 6–15 Gravel (G-W) 47–95 2.7–5 Note: The table was modified by the authors of this document so that the DPI and CBR correlation matched. Typical Results Burnham (1997) described an extensive set of DCP measurements on the subgrade soils and base materials used in the various test sections at the MnRoad facility. These are summarized in Table 1.25. Following this work, the following DPI limits were recommended for use by MnDOT personnel when analyzing DCP results for rehabilita tion studies: silty/clay materials, DPI <25 mm/blow; select granular mate- rials, DPI <7 mm/blow; and Class 3 special gradation materials, DPI <5 mm/blow. Subgrade Stability The Illinois DOT (1982, 2005) has used the DCP to check subgrade stability. The purpose of this is straightforward—they want to know if the subgrade is stable enough to avoid excessive rutting and/or shoving during and following construction activities. The subgrade immediate bearing value (IBV) can be estimated from the DPI. The IBV is similar to the CBR, “except that IBV testing is conducted on a 4-inch molded sample instead of the CBR’s 6-inch sample . . . further, the penetration test for determining the IBV is conducted immediately after compaction instead of waiting 96 hours—thus IBV and CBR are similar but not identical” (Illinois DOT, 2005). Figure 1.30 shows the relationship between unsoaked CBR (actually IBV), DPI, and required thickness of

47 PROJECT ASSESSMENT MANUAL TABLE 1.25. MINNESOTA DCP RESULTS FOLLOWING PLACEMENT OF THE BASE COURSE Material DPI Average (mm/blow) (SD) 0–12 in. depth DPI Average (mm/blow) (SD) 12–24 in. depth DPI Average (mm/blow) (SD) 24–36 in. depth Clay/Silt Location 1 11 (3) 21 (7) 21 (7) Clay/Silt Location 2 14 (6) 18 (5) 16 (5) Clay/Silt Location 3 12 (5) 20 (7) 15 (7) Sand 5 (2) 5 (1) 6 (2) Base Course 4 (2) 3 (1) 3 (<1) Note: DPI average values were rounded to the nearest whole number. Figure 1.30. DCP-based thickness design for granular backfill and subgrade modification for the Illinois DOT. Source: Burnham, 1997; checked against Illinois DOT, 2005.

48 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE remedial measures. Remedial measures can include the addition of granular backfill or subgrade modification such as lime stabilization. The Illinois DOT DCP results and those from the Minnesota DOT broadly agree in that subgrade DPI values greater than 25 mm/blow are of concern. Use of DCP Data in Renewal Decisions The Texas Transportation Institute developed guidelines for the Texas DOT as to con- ditions suitable for rubblizing existing rigid pavements (Figure 1.31). The “High Risk” portion of the figure implies the pavement is not a good candidate for rubblization because the supporting base and subgrade is excessively weak. Figure 1.31 is similar to, but modified from, similar guidelines developed for Illinois (Figure 1.32). Figure 1.32 is of interest because it includes data obtained by Sebesta and Scullion (2007) for US-83 in Texas plotted by total pavement thickness versus DCP-derived CBR values. Figure 1.31. Rubblization selection chart developed by TTI. Source: Sebesta and Scullion, 2007.

49 PROJECT ASSESSMENT MANUAL SUBGRADE SOIL SAMPLING AND TESTS Purpose This section overviews selected elements associated with subgrade soils and what in- formation is needed to make pavement assessment decisions. Much of pavement anal- ysis and understanding stems from knowledge of layer thicknesses, types of materials, and condition. Measurement Methods This subsection shows both the types of tests and the frequency of sampling associated with subgrade soils. A summary of these tests is contained in Table 1.26. Figure 1.32. Illinois rubblization selection chart with data from US-83 (Texas). Source: Sebesta and Scullion, 2007; original Illinois DOT criteria from Heckel, 2002.

50 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Analysis Tools Questions that need to be answered for the project assessment regarding subgrade soils include the following: • How well do the subgrade soils support the existing pavement structure? • Are the subgrade soils frost susceptible (if the project is located within a potential freezing zone)? • Are the subgrade soils subject to expansion and contraction (such as expansive clay soils)? • Are groundwater issues associated with the project site? Support for Existing Pavement Structure The support for the existing pavement structure can be estimated through a combina- tion of laboratory or nondestructive testing—but most likely it will be NDT. A set of FWD deflection basins, pavement coring, and DCP measurements is generally suffi- cient, along with use of the analysis tools provided in the preceding sections. Frost Susceptibility Both sophisticated and very straightforward soils tests are available for estimating the likelihood of subgrade soil frost susceptibility. The basic issue is the potential for the creation of ice lenses under the existing pavement and the resulting loss of support TABLE 1.26. SUMMARY OF TYPICAL SUBGRADE TESTS Subgrade Test Standard Test Method Purpose of Test Soil classification ASTM D2487-00, Standard Classification of Soils for Engineering Purposes (Unified Soil Classification System) Soil classification is basic information that can be used to estimate various design-related parameters. The required tests for classification can be used for other determinations (gradation, Atterberg limits). California bearing ratio ASTM D1883-07e2, Standard Test Method for CBR of Laboratory-Compacted Soils Straightforward test for determining relative shear strength of the subgrade soils. CBR can be estimated from a laboratory test or through correlations with devices such as the DCP (see section “Dynamic Cone Penetrometer (DCP)” in this chapter). Caution is needed because laboratory- and field-produced CBRs can have quite different moisture conditions and, hence, different results. Resilient modulus— laboratory AASHTO T307, Standard Method of Test for Determining the Resilient Modulus of Soils and Aggregate Materials If subgrade soil samples are available, laboratory resilient modulus determinations can be made. Triaxial testing is expensive and the results are a function of sample preparation. Resilient modulus—NDT ASTM D4694-96, Standard Test Method for Deflections with a Falling-Weight-Type Impulse Load Device The preferred test apparatus for nondestructive testing of pavement structures is the FWD (see section “Nondestructive Testing via the Falling Weight Deflectometer (FWD)” in this chapter). Straightforward methods for estimating MR are available (same section), or backcalculation procedures allow up to three pavement layers to be estimated.

51 PROJECT ASSESSMENT MANUAL when it all thaws out. When ice lenses form in frost-susceptible soils, large volume changes can occur (just liquid water changing to ice increases the volume by 9%). An illustration of ice lenses in pavements in shown in Figure 1.33. A basic approach for assessing frost susceptibility is based on gradation, and it has been in use for almost 80 years. Casagrande noted the following in 1932 (taken from Terzaghi and Peck, 1967): “Under natural freezing conditions and with sufficient water supply one should expect considerable ice segregation in non-uniform soils con- taining more than 3% of grains smaller than 0.02 mm. . . . No ice segregation was observed in soils containing less than 1% of grains smaller than 0.02 mm, even if the groundwater level is as high as the frost line.” To determine the percent passing 0.02 mm requires a hydrometer test. A reasonable approximation of 3% passing 0.02 mm is about 7% passing a 0.075 mm (No. 200 sieve). Figure 1.33. Formation of ice lenses in a pavement structure.

52 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Another tool that can aid decisions about the potential frost susceptibility of a subgrade soil is to use the U.S. Army Corps of Engineers classification system for frost design, as shown in Table 1.27. Expansion and Contraction If these types of soils are present, attempt to answer the following: • Were the subgrade soils previously treated with materials such as lime? • Is the profile of the existing pavement stable? Groundwater Issues When groundwater issues are apparent, investigation by a geotechnical engineer may be required. TABLE 1.27. U.S. ARMY CORPS OF ENGINEERS FROST DESIGN SOIL CLASSIFICATION Frost Group Soil Type Percentage Finer than 0.02 mm by Weight (%) Typical Soil Types under Unified Soil Classification System Nonfrost susceptible (NFS) (a) Gravels, including crushed stone and crushed rock 0–1.5 GW, GP (b) Sands 0–3 SW, SP Potentially frost susceptible (PFS) (a) Gravels Crushed stone Crushed rock 1.5–3 GW, GP (b) Sands 3–10 SW, SP S1 Gravelly soils 3–6 GW, GP, GW-GM, GP-GM S2 Sandy soils 3–6 SW, SP, SW-SM, SP-SM F1 Gravelly soils 6–10 GM, GW-GM, GP-GM F2 (a) Gravelly soils 10–20 GM, GW-GM, GP-GM (b) Sands 6–15 SM, SW-SM, SP-SM F3 (a) Gravelly soils >20 GM, GC (b) Sands, except very fine silty sands >15 SM, SC (c) Clays, PI >12 — CL, CH F4 (a) All silts — ML, MH (b) Very fine silty sands >15 SM (c) Clays, PI <12 — CL, CL-ML (d) Varved clays and other fine- grained, banded sediments — CL and ML; CL, ML, and SM; CL, CH, and ML; CL, CH, ML, and SM Note: Table after U.S. Army, 1990, and NCHRP Synthesis 26, 1974.

53 PROJECT ASSESSMENT MANUAL TRAFFIC LOADS FOR DESIGN Purpose This section overviews the use of basic traffic information to estimate loadings for pavement design. The fundamental parameter to be estimated is the equivalent single axle load (ESAL). More detailed assessments of traffic loading such as load spectra used in the Mechanistic-Empirical Pavement Design Guide (MEPDG) are not needed for use in these guidelines. Measurement Method This subsection overviews the kind of traffic information needed to quickly estimate future ESALs. Tire Loads and Terminology Typical truck and bus axles are shown in Figure 1.34, which illustrates single and tan- dem axles with either single or dual tires. States generally have regulations limiting the allowable load per inch of tire width. This tire load limitation varies from a high of 800 lb/in. to a low of 450 lb/in. The primary impact of such state laws has to do with the use of dual or single tires on a specific axle and steer axles. Figure 1.34. Illustration of typical axle and tire configurations.

54 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Typical Federal and State Axle Load Limits Typical federal and state axle load limits are as follows: • Single axles, 20,000 lb; • Tandem axles, 34,000 lb; and • Total truck gross weight, 80,000 lb. FHWA Bridge Formula A major additional limitation on U.S. trucks and buses is the FHWA bridge formula. The total gross weight in pounds imposed on the pavement by any group of two or more consecutive axles on a vehicle or combination of vehicles shall not exceed that weight calculated by use of Equation 1.10 below. The bridge formula is needed since an individual set of bridge design computations cannot be performed for every type of truck that may use highways. Bridge designers use a standard design vehicle for estimating critical stresses, strains, or deflections in a bridge structure. This vehicle is designated HS-20-44 and has been referred to as an umbrella loading. Federal law requires its use in bridge design for the Interstate system. In effect, the bridge formula helps to ensure bridges are not “overstressed” due to an almost infinite number of truck-axle configurations and weights. The equation is W = 500(NL/(N−1) + 12N + 36) (1.10) where W = maximum weight on any group of two or more consecutive axles to the near- est 500 lb, L = distance between the extremes of any group of two or more consecutive axles (ft), and N = number of axles in the group under consideration. To illustrate, an example is a five-axle truck with a 51-ft separation from the steer axle to the rear portion of the back tandem. The total vehicle allowable gross weight, via the bridge formula, is then W = 500(5(51)/(5−1) + 12(5) + 36) = 80,000 lb. Repetitions of Wheel Loads and Equivalent Single Axle Loads (ESALs) To compute ESALs, we must be able to convert wheel loads of various magnitudes and repetitions (“mixed traffic”) to an equivalent number of “standard” or ”equivalent” loads for design purposes. The most commonly used equivalent load is 18,000 lb (80 kN) equivalent single axle loads (normally designated ESAL). The ESAL standard axle load is used in the AASHTO “Guide for Design of Pavement Structures” (AASHTO, 1993).

55 PROJECT ASSESSMENT MANUAL Wheel load equivalency has been one of the most widely adopted results of the AASHO Road Test (1958 to 1960) and has provided a method to relate relative dam- age attributed to axles of different type (single and tandem) and weight. Highway design in most states is based on the ESAL traffic input anticipated over a future 10- to 50-year period. The relationship between repetitions is not arithmetically proportional to the axle loading. Instead, a 10,000-lb single axle needs to be applied to a pavement struc- ture many more than 1.8 times the number of repetitions of an 18,000-lb single axle to have the same effect; in fact, it must be applied more than 12 times. Similarly, a 22,000-lb single axle needs to be repeated less than half the number of times of an 18,000-lb single axle to have an equivalent effect. A sample of ESAL load equivalency factors (LEFs) is shown in Table 1.28. A basic element in estimating the future ESALs for a specific project is to forecast the truck and bus volumes for the design (and analysis) period. Once this is done, LEFs in various forms can be applied to the forecast volumes and summed. A complete forecast will include the 13 FHWA vehicle classes (which are not the same vehicle classes as those used by vehicle manufacturers). These classes are shown in Table 1.29. A somewhat simplified scheme for summarizing the 13 vehicle classes in Table 1.29 is to group all truck and bus traffic into three groups or units as shown in Table 1.30. TABLE 1.28 SAMPLE OF AASHTO EQUIVALENCY FACTORS Axle Type (lb) Axle Load (lb) ESAL Load Equivalency Factorsa Single axle 2,000 10,000 14,000 18,000 20,000 30,000 0.0003 0.118 0.399 1.000 1.4 7.9 Tandem axle 2,000 10,000 14,000 18,000 20,000 30,000 34,000 40,000 50,000 0.0001 0.011 0.042 0.109 0.162 0.703 1.11 2.06 5.03 a Data from AASHTO (1993).

56 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 1.29. FHWA VEHICLE CLASSES FHWA Vehicle Class Vehicle Class Description Class 1 Motorcycles (Optional): All two- or three-wheeled motorized vehicles. Typical vehicles in this category have saddle-type seats and are steered by handlebars rather than wheels. This category includes motorcycles, motor scooters, mopeds, motor-powered bicycles, and three-wheel motorcycles. This vehicle type may be reported at the option of the state. Class 2 Passenger Cars: All sedans, coupes, and station wagons manufactured primarily for the purpose of carrying passengers and including those passenger cars pulling recreational or other light trailers. Class 3 Other Two-Axle, Four-Tire Single-Unit Vehicles: All two-axle, four-tire vehicles, other than passenger cars. Included in this classification are pickups, panels, vans, and other vehicles such as campers, motor homes, ambulances, hearses, and carryalls. Other two-axle, four-tire single-unit vehicles pulling recreational or other light trailers are included in this classification. Class 4 Buses: All vehicles manufactured as traditional passenger-carrying buses with two axles and six tires or three or more axles. This category includes only traditional buses (including school buses) functioning as passenger-carrying vehicles. All two-axle, four-tire single-unit vehicles. Modified buses should be considered to be a truck and be appropriately classified. Class 5 Two-Axle, Six-Tire Single-Unit Trucks: All vehicles on a single frame, including trucks, camping and recreational vehicles, and motor homes, having two axles and dual rear wheels. Class 6 Three-Axle Single-Unit Trucks: All vehicles on a single frame, including trucks, camping and recreational vehicles, motor homes, having three axles. Class 7 Four or More Axle Single-Unit Trucks: All trucks on a single frame with four or more axles. Class 8 Four or Fewer Axle Single-Trailer Trucks: All vehicles with four or fewer axles consisting of two units, one of which is a tractor or straight truck power unit. Class 9 Five-Axle Single-Trailer Trucks: All five-axle vehicles consisting of two units, one of which is a tractor or straight truck power unit. Class 10 Six or More Axle Single-Trailer Trucks: All vehicles with six or more axles consisting of two units, one of which is a tractor or straight truck power unit. Class 11 Five or Fewer Axle Multi-Trailer Trucks: All vehicles with five or fewer axles consisting of three or more units, one of which is a tractor or straight truck power unit. Class 12 Six-Axle Multi-Trailer Trucks: All six-axle vehicles consisting of three or more units, one of which is a tractor or straight truck power unit. Class 13 Seven or More Axle Multi-Trailer Trucks: All vehicles with seven or more axles consisting of three or more units, one of which is a tractor or straight truck power unit. TABLE 1.30. SIMPLIFIED TRUCK AND BUS GROUPS Simplified Vehicle Categories Groupings of FHWA Vehicle Classes Single Units (i) Buses (FHWA Class 4) (ii) 2-axle, 6-tire single units (FHWA Class 5) (iii) 3-axle single units (FHWA Class 6) (iv) 4+-axle single units (FHWA Class 7) Single Trailers (i) 4-axle single trailer (FHWA Class 8) (ii) 5-axle single trailer (FHWA Class 9) (iii) 6+-axle single trailer (FHWA Class 10) Multi-Trailers (i) 5-axle multi-trailer (FHWA Class 11) (ii) 6-axle multi-trailer (FHWA Class 12) (iii) 7+-axle multi-trailer (FHWA Class 13)

57 PROJECT ASSESSMENT MANUAL Analysis Tools This subsection focuses on how to organize ESAL data so that an overall ESAL esti- mate for the design period can be made. Table 1.31 shows typical ESALs per vehicle according to the groupings in Table 1.30. The ESALs per vehicle were developed by a state DOT and appear to be typi- cal for U.S. truck traffic. They may appear to be low, but the values are averages that include empty backhauls. TABLE 1.31. ESALS PER VEHICLE FOR SIMPLIFIED VEHICLE GROUPS Simplified Vehicle Categories FHWA Classes Average ESALs per Vehicle Single-Unit Trucks 4, 5, 6, 7 0.40 Trucks with Single Trailers 8, 9, 10 1.00 Trucks with Multi-Trailers 11, 12, 13 1.75 Buses (half full) 4 1.60 Thus, if you estimated that a specific highway has daily (one-way) 1,000 single- unit trucks, 2,000 trucks with single trailers, and 500 trucks with multi-trailers and no buses, then the daily ESALs would be [1,000(0.4) + 2,000(1.00) + 500(1.75)] = 3,275 ESALs per day or about 1,200,000 ESALs per year. The annual value can be scaled up to the design period with a suitable growth rate (typically 2% to 3%). CONSTRUCTION PRODUCTIVITY AND TRAFFIC IMPACTS Purpose This section overviews the various methods for determining construction productivity and traffic impacts of pavement and roadway construction. Traffic impacts can often make up the largest societal cost associated with a paving project, sometimes being an order of magnitude more than the agency cost to build or rehabilitate the pavement. An early understanding of productivity and potential traffic impacts can assist the project in determining the most advantageous construction timing, project sequenc- ing (staging), and lane-closure scenarios. Often, full roadway closures (in contrast to repeated partial closures) over longer periods of time (e.g., full weekends or multiple days instead of nighttime-only closures) can prove to be the least costly alternative if user costs are properly accounted for in construction planning. Measurement Methods Traffic impacts are typically quantified by user delay, with typical metrics being (1) total user delay, (2) total user cost associated with delay, (3) maximum vehicle queue length, and (4) maximum time in vehicle queue. Usually, the goal of minimizing traffic impacts is interpreted to mean minimizing the total user cost attributable to the existence of the project work zone. Other important considerations (e.g., accident and incident

58 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE minimization, avoidance of certain public event days that generate high traffic) may cause the ultimate traffic impacts to be somewhat greater than the optimal minimum. Nonetheless, it is useful to estimate, as accurately as practical considerations allow, the minimum traffic impact scenario for pavement construction. Generally, this estimate uses the following six basic actions: 1. Determine construction productivity. This involves estimating the productivity of basic construction processes associated with the project such as demolition crew speed and efficiency, dump truck number and capacity, paver speed, and mate- rials manufacturing plant productivity. It also involves estimating mobilization and demobilization times, concrete cure time, hot-mix asphalt cooling time, and traffic control setup time. There may be several estimates of each depending on the construction scenarios being investigated. 2. Measure existing traffic. Although an actual time history is best (e.g., from loop detector information or manual counts), average daily traffic (ADT) can be used and hourly traffic volumes can be developed by multiplying ADT by typical hourly distribution factors for the type of roadway being analyzed. 3. Estimate the fraction of traffic that will cancel trips and the fraction of traffic that will use detour routes during the construction. At best, these will be rough estimates unless extremely sophisticated models are used. These estimates are also highly dependent on the publicity given the roadway work. Values can be obtained from (a) agency experience with similar closures and similar publicity in the past and (b) a general literature review of similar traffic closures. 4. Develop construction scheduling (staging) alternatives. This involves determin- ing the number, duration, and sequence of lane closures required to complete the project. As the traffic impact analysis progresses, it is often necessary to refine these alternatives. Strong consideration should be given to scheduling alternatives that result in work-zone traffic capacity greater than traffic demand during the hours of work. Essentially, this results in little or no user cost attributed to the roadway work. However, such scheduling alternatives may not exist or be feasible from a construction productivity and/or constructability standpoint. Any number of lane-closure scenarios can be considered, but it is helpful to at least investigate the follow ing six scenarios. a. Partial night closures involve closure only during night hours with light traffic where each roadway direction is still open, although with reduced capacity in at least one direction. These closures are often the first considered because they tend to minimize traffic impacts by only closing lanes when traffic is the lightest. However, they may not provide the lowest user costs because mobilization and demobilization can take up a large percentage of total closure time, resulting in low overall productivity. In some scenarios, it may not be possible to make any meaningful progress in a short nighttime closure. Even if partial night closures cannot be used for mainline paving, they are often useful for prepaving work (e.g., PCC panel saw-cutting, restriping lanes, milling HMA).

59 PROJECT ASSESSMENT MANUAL b. Full night closures are the same as above but with at least one roadway direc- tion fully closed. These may involve detouring an entire direction, counterflow- ing traffic on one side of a highway, or using a pilot car to alternate traffic direc- tions in one lane. Full night closures sometimes require such things as setting up counterflow traffic on one side of a roadway or accomplishing dangerous overhead work such as overpass demolition or placement. c. Partial day closures are closures only during day hours where each roadway direction is still open, although with reduced capacity in at least one direc- tion. These closures are often the first considered for lightly trafficked roadways where user delay is unexpected even with some lanes closed. If traffic delays are minimal, day closures can improve safety by providing better visibility and en- countering fewer impaired drivers than night work, and can reduce construction costs by avoiding overtime pay. However, they may not provide the lowest user costs because mobilization and demobilization can take up a large percentage of total closure time, resulting in low overall productivity. In some scenarios, it may not be possible to make any meaningful progress in a partial day closure. d. Full day closures are the same as above but with at least one roadway direction fully closed. These may involve detouring an entire direction, counterflowing traffic on one side of a highway, or using a pilot car to alternate traffic direc- tions in one lane. Full day closures are usually only feasible for lightly trafficked roadways or roadways with large-capacity detour routes that do not add sig- nificantly to commute time. e. Partial or full weekend continuous closures start Friday evening after peak hour traffic and end Monday morning before peak hour traffic. The typical scenario is a 55-hour weekend closure starting at 9 or 10 p.m. on Friday night and end- ing at 4 or 5 a.m. on Monday morning. The long closure time allows for better productivity because mobilization and demobilization takes up a smaller frac- tion of total closure time and, more importantly, because construction crews generally get better and faster in their work given a longer working window. Weekends are typically preferred because weekend traffic is usually more dis- cretionary (leading to more canceled trips and less total user delay) and often lighter than weekday traffic. f. Partial or full week-long continuous closures are maintained continuously over an entire week (168 hours). Although it may not be known if any closure win- dows will extend over a week or more, estimating this alternative will gener- ally allow estimation of longer closure windows with reasonable accuracy. For instance, the productivity for a 3-week closure is roughly, but not exactly (due to mobilization and demobilization times), three times the productivity of a 1-week continuous closure.

60 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE 5. Model traffic using the tool of choice (see the section “Analysis Tools”). This modeling will result in an estimate of total user cost for the roadway project. In general, larger projects on major routes warrant more modeling, whereas smaller projects on minor routes can often be estimated sufficiently using spreadsheets. 6. Apply FHWA Interim Report Life-Cycle Cost Analysis in Pavement Design (Walls and Smith, 1998) standards to estimate user delay cost. This report provides rea- sonable values for user time (Table 1.32). The values in this table are in 1996 dollars and should be adjusted to current dollars using the Consumer Price Index. A simple calculator is available from the U.S. Bureau of Labor Statistics at http:// www.bls.gov/data/inflation_calculator.htm. Multiplying these values by total delay for each class of vehicle gives an estimate of total work zone user delay cost. Additionally, FHWA has accumulated additional information that can be found at http://www.fhwa.dot.gov/infrastructure/asstmgmt/lcca.cfm. TABLE 1.32. RECOMMENDED VALUES OF TIME Vehicle Class Value per Vehicle Hour (1996 dollars) Value Range Passenger Vehicles $11.58 $10–$13 Single-Unit Trucks $18.54 $17–$20 Combination Trucks $22.31 $21–$24 Note: From Walls and Smith, 1998. General Guidance The following general guidance for traffic impacts comes largely from the guidance documents listed in the references for this section. Closure Scenarios • Productivity is usually much higher and worker safety is greater with longer, more complete closures (e.g., full closures, weekend closures; FHWA, 2003). • The public is generally very accepting of full closures or a few longer-duration closures as an alternative to lengthy schedules of night or day closures (FHWA, 2003). • As a work zone remains in effect for a longer period of time (e.g., over several days or several weekends), the fraction of drivers either canceling their trips or taking the detour route is likely to decrease as drivers become used to the situation or determine that a trip can no longer be put off. • Detour routes may experience several times their normal traffic volumes (Lee, Lee, and Harvey, 2006; Lee et al., 2001). It may be prudent to improve detour route capacity through additional lanes, a temporarily reversible lane, signal retiming, or other improvements (FHWA, 2003).

61 PROJECT ASSESSMENT MANUAL • For major highway jobs, the construction of one lane usually requires a second adjacent lane for access. This means either using an existing wide shoulder (e.g., 10-ft shoulder) if one exists or closing a second lane (Lee, 2008). • For major highway jobs, if the lane under construction has more than one major activity under way on it simultaneously (e.g., demolition and paving), a second access lane will likely be needed to avoid stationary trucks in the adjacent lane (Lee, 2008). • Avoid creating work zones with live traffic on both sides (e.g., in the middle lanes in one direction). These generally do not leave workers a safe exit from the work zone if it is compromised. • It may be better to use a simpler lane-closure plan that is more easily understood by the public even if it does not result in the minimum modeled user delay. Contracting • Lane rental or time-based bonus/penalty contracts should have a clear clause de- scribing how to address changed conditions or any situation where the owner wishes to add work that impacts productivity (Lee et al., 2007). Often, contrac- tors plan to spend more money than the contract price in order to finish early and receive the bonus. In this scenario, without bonus payments, the contractor will lose money. • Contracts that contain bonus/penalty amounts for speed and quality should balance these amounts so that it does not become advantageous to sacrifice one bonus to get the other (Muench et al., 2007). For instance, of a maximum quality bonus/penalty is $3,000 but the maximum speed bonus/penalty is $100,000 then in some scenarios it may be logical to sacrifice a small quality bonus for a large speed bonus. Productivity • The slowest process in a reconstruction project is often demolition (Lee, Lee, and Ibbs, 2007). If several processes are being done simultaneously, demolition will most often control the overall productivity. • The rate at which dump trucks can be filled by an excavator or milling machine is relatively consistent from job to job (Lee et al., 2007). Therefore, the best estimate is often what happened on the previous job. If no local information is available, Lee et al. (2007) provide good baseline estimates. • Production rate is often controlled by access to the construction site and allow- ances made for traffic (e.g., temporary off-ramps in work zones, separation be- tween work zone and traffic).

62 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Work-Zone Capacity • Work-zone capacity is highly variable and only moderately predictable. Work-zone capacity can be affected by the number of lanes open, intensity of work, presence of ramps, fraction of heavy vehicles, lane width, lateral clearance, work-zone grade, and more. Highway Capacity Manual (HCM) (TRB, 2000) procedures are very rough, but they suggest 1,600 passenger cars per lane per hour (pc/lane/h) be used as a baseline for short-term work zones. Typically this number is adjusted down- ward based on other factors and can be as low as about half the original value. • The more a work zone can be physically and visually separated from traffic (e.g., semipermanent barriers like jersey barriers or k-rails instead of traffic cones or barrels), the greater the work-zone traffic capacity. • Incidents (i.e., accidents, stalled vehicles, etc.) are one of the largest contributors to work zone user delay because there are fewer lanes (if any) that traffic can use to bypass the incident. Dedicating resources (e.g., incident response vehicle, video cameras, variable message boards, traffic management center) to reduce incidents and clear them more quickly can be a cost-effective way to minimize user delay (FHWA, 2004). Publicity • Roadway work and closure publicity can be effective in drastically reducing traffic during work-zone closures. Often, several-mile-long queues predicted using nor- mal traffic volumes never materialize because many drivers cancel their trips or alter their routes. • Even if a local public information campaign is effective, it may still be difficult to get closure information to travelers or freight carriers out of the local area who plan on using the affected roadway. Analysis Tools This subsection overviews some of the more popular methods for determining traffic impacts for pavement construction projects and factors that influence the choice of tools. Some key considerations when selecting tools are as follows: • How much detail is needed? Work-zone characteristics, desired outputs, and the stage of planning, design, and construction will influence tool choice. Often a sim- pler tool, with less detail, is adequate. • Is the tool calibrated to the local area? If not, results may still be useful; however, accuracy may be less than expected or needed. • Is the tool stochastic or deterministic? Construction productivity and traffic can be highly variable and difficult to predict. Although a deterministic model can provide a single number, it is better to provide a reasonable range of answers to capture the variable nature of productivity and traffic.

63 PROJECT ASSESSMENT MANUAL • How much detail does the tool produce? Some tools can only estimate traffic im- pacts over one 24-hour period while others can estimate over much longer time periods. Some tools can only estimate delay on an hourly basis, whereas others can estimate them in much smaller time increments. Some tools make estimates using one single day’s traffic input, whereas others are able to account for daily, weekly, and monthly traffic variations. Analysis Tools: Construction Productivity Construction productivity tools discussed are manual methods, standard estimating software, and Construction Analysis for Pavement Rehabilitation Strategies (CA4PRS). Manual method. Demolition and paving productivity estimates can be made manually by comparing productivities of the constituent processes and identifying the limiting factor. There are a few references to help in paving productivity calculations. The National Asphalt Pavement Association (NAPA) publishes Balancing Production Rates in Hot Mix Asphalt Operations (IS 120), which contains a step-by-step guide for determining HMA paving productivity. Several companies also offer custom-printed asphalt productivity slide rules that paving companies can purchase and brand to be given out to potential customers. Estimating software. Most estimating software (e.g., Bid2Win, HeavyBid) assists users in calculating the productivity of construction processes. CA4PRS. CA4PRS is a Microsoft Access–based software tool that can be used to analyze highway pavement rehabilitation strategies including productivity, project scheduling, traffic impacts, and initial project costs based on input data and constraints supplied by the user. The goal is to help determine roadway rehabilitation strategies that maximize production and minimize costs without creating unacceptable traffic delays. As of 2009, all state transportation departments have free group licenses for CA4PRS. First-Order Productivity Estimates In the early planning stages of a project, it may be useful to quickly determine rough construction productivity based on a few known parameters. This section displays productivity graphs produced using CA4PRS with most inputs being held constant at typical values. The purpose of these graphs is only to give a rough estimate of typical productivity. CA4PRS should be used to produce more accurate numbers based on actual site-specific parameters for use in any project planning. In general, most inputs were fixed except for the trucking rates (i.e., removal of demolition from the site and delivery of paving material to the site). Thus, the 95% confidence intervals seen are mostly dependent on these delivery rates. In all cases, a 10-mi stretch of two lanes was analyzed (20 lane miles total). As with all data input values, this length of highway and total lane miles have some influence on productivity. Tables 1.33 through 1.38 show input parameters used in CA4PRS to generate Figures 1.35 through 1.43. Estimates are given for the following:

64 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 1.33. CA4PRS INPUT VALUES FOR REMOVE AND REPLACE WITH PCC Input Value Distribution/Comments Activity Constraints Mobilization 1.0 h None, deterministic Demobilization 2.0 h None, deterministic Base paving None NA Demo-to-PCC paving lag times for sequential method 1.0 h Triangular (min = 0.5 h, max = 1.5 h) Demo-to-PCC paving lag times for concurrent method 2.0 h Triangular (min = 1.0 h, max = 3.0 h) Resource Profile Demolition Hauling Truck Rated capacity 18.0 tons 9 yd3 of a 15-yd3 truck filled with 2.0 tons/yd3 material Trucks/h/team 10 trucks Triangular (min = 8 trucks, max = 12 trucks) Packing efficiency 1.0 None, deterministic Number of teams 1.0 2.0 1 team for screed paving, 2 teams for slipform None, deterministic Team efficiency 0.90 Triangular (min = 0.85, max = 0.95) Base Delivery Truck None NA (no base material) Batch Plant Capacity 500 yd3/h None, deterministic (set high to ensure plant is not the limiting activity) Number of plants 1 None, deterministic Concrete Delivery Truck Capacity 7.5 yd3 NA Trucks per hour 10/h 13/h The first rate is for screed paving and the second is for slipform paving Triangular (min = 8/h, max = 12/h) Triangular (min = 15/h, max = 19/h) Packing efficiency 1.0 None, deterministic Paver Speed 5 ft/min None, deterministic Number of pavers 1 None, deterministic Schedule Analysis Construction window See graphs Section profile See graphs Note: No base material included in graphs Change in roadway elevation No change Lane widths 12 ft Curing time 12 h Working method See graphs

65 PROJECT ASSESSMENT MANUAL TABLE 1.34. CA4PRS INPUT VALUES FOR REMOVE AND REPLACE WITH HMA Input Value Distribution/Comments Activity Constraints Mobilization 1.0 h None, deterministic Demobilization 2.0 h None, deterministic Base paving None NA Demo-to-HMA paving lag 1.0 h Triangular (min = 0.5 h, max = 1.5 h) Half closure traffic switch 0.5 h Triangular (min = 0.25 h, max = 0.75 h) Resource Profile Demolition Hauling Truck Rated capacity 18.0 tons 9 yd3 of a 15-yd3 truck filled with 2.0 tons/yd3 material Trucks/h/team 10 trucks Triangular (min = 8 trucks, max = 12 trucks) Packing efficiency 1.0 None, deterministic Number of teams 1.0 None, deterministic Team efficiency 0.90 Triangular (min = 0.85, max = 0.95) Paver None NA (no base material) Nonpaving speed 15 mph Batch Plant Capacity 500 yd3/h None, deterministic (set high to ensure plant is not the limiting activity) Number of plants 1 None, deterministic HMA Delivery Truck Capacity 18 tons NA Trucks per hour 12/h Triangular (min = 10/h, max = 14/h) Packing efficiency 1.0 None, deterministic Schedule Analysis Construction window See graphs Section profile See graphs Top two lifts are 2 in. each, all other lifts are 3 in. each; paver moves at 0.6 mph for top two lifts and 0.5 mph for all other lifts Change in roadway elevation No change Shoulder overlay Prepaving Shoulder overlays are not accounted for Curing time 12 h Working method See graphs Cooling time analysis User specifications Time calculated in MultiCool and manually entered Lane Widths Number of lanes 2 Lane widths 12 ft each

66 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 1.35. CA4PRS INPUT VALUES FOR MILL AND FILL WITH HMA Input Value Distribution/Comments Activity Constraints Mobilization 1.0 h None, deterministic Demobilization 2.0 h None, deterministic Mill-to-HMA paving lag 1.0 h Triangular (min = 0.5 h, max = 1.5 h) Half closure traffic switch 0.5 h Triangular (min = 0.25 h, max = 0.75 h) Resource Profile Milling and Hauling Number of teams 1.0 None, deterministic Team efficiency 0.90 Triangular (min = 0.85, max = 0.95) Milling Machine Class Large Material type AC-Hard Efficiency factor 0.90 Triangular (min = 0.85, max = 0.95) Hauling Truck Rated capacity 18.0 tons 9 yd3 of a 15-yd3 truck filled with 2.0 tons/yd3 material Trucks/h/team 13 trucks Triangular (min = 11 trucks, max = 15 trucks) Packing efficiency 1.0 None, deterministic Batch Plant Capacity 500 yd3/h None, deterministic (set high to ensure plant is not the limiting activity) Number of plants 1 None, deterministic HMA Delivery Truck Capacity 18 tons NA Trucks per hour 12/h Triangular (min = 10/h, max = 14/h) Packing efficiency 1.0 None, deterministic Paver None NA (no base material) Nonpaving speed 15 mph Schedule Analysis Construction window See graphs Section profile See graphs Lifts are between 1.5 and 3 in.; paver speeds are 0.5–0.6 mph Change in roadway elevation No change Shoulder overlay Prepaving Shoulder overlays are not accounted for Curing time 12 h Working method See graphs Cooling time analysis User specifications Time calculated in MultiCool and manually entered Lane Widths Number of lanes 2 Lane widths 12 ft each

67 PROJECT ASSESSMENT MANUAL TABLE 1.36. CA4PRS INPUT VALUES FOR CRACK, SEAT, AND OVERLAY Input Value Distribution/Comments Activity Constraints Mobilization 3.0 h None, deterministic Demobilization 2.0 h None, deterministic Half closure traffic switch 0.5 h Triangular (min = 0.25 h, max = 0.75 h) Resource Profile Paver None NA (no base material) Nonpaving speed 15 mph Batch Plant Capacity 500 yd3/h None, deterministic (set high to ensure plant is not the limiting activity) Number of plants 1 None, deterministic HMA Delivery Truck Capacity 18 tons NA Trucks per hour 12/h Triangular (min = 10/h, max = 14/h) Packing efficiency 1.0 None, deterministic Schedule Analysis Construction window See graphs Section profile See graphs Top two lifts are 2 in. each, all other lifts are 3 in. each; paver moves at 0.6 mph for top two lifts and 0.5 mph for all other lifts Change in roadway elevation No change Shoulder overlay Prepaving Shoulder overlays are not accounted for Curing time 12 h Working method See graphs Cooling time analysis User specifications Time calculated in MultiCool and manually entered Lane Widths Number of Lanes 2 Lane widths 12 ft each

68 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE TABLE 1.37. CA4PRS INPUT VALUES FOR UNBONDED PCC OVERLAY Input Value Distribution/Comments Activity Constraints Mobilization 3.0 h None, deterministic (longer time accounts for surface preparation) Demobilization 2.0 h None, deterministic Base paving None NA Demo-to-PCC paving lag times for sequential method 0 h No demolition occurs Demo-to-PCC paving lag times for concurrent method 0 h No demolition occurs Resource Profile Demolition Hauling Truck High numbers are a work-around to make demolition take essentially no time Rated capacity 100.0 tons None, deterministic Trucks/h/team 100 trucks None, deterministic Packing efficiency 1.0 None, deterministic Number of teams 100.0 None, deterministic Team efficiency 1.00 None, deterministic Base Delivery Truck None NA (no base material) Batch Plant Capacity 500 yd3/h None, deterministic (set high to ensure plant is not the limiting activity) Number of plants 1 None, deterministic Concrete Delivery Truck Capacity 7.5 yd3 NA Trucks per hour 10/h Triangular (min = 8/h, max = 12/h) Packing efficiency 1.0 None, deterministic Paver Speed 5 ft/min None, deterministic Number of pavers 1 None, deterministic Schedule Analysis Construction window See graphs Section profile See graphs Note: No base material included in graphs Change in roadway elevation No change Lane widths 12 ft Curing time 12 h Working method See graphs

69 PROJECT ASSESSMENT MANUAL • Remove and replace with PCC. Remove the existing pavement and replace with the same depth of new PCC pavement. Productivity is estimated for sequential operations (only one major operation—demolition or paving—is occurring on the jobsite at any one time) and concurrent operations (both major operations— demolition and paving—are occurring on the jobsite at once, with the appropriate space in between). One lane is paved at a time. Sequential operations require one additional lane shut down for construction access, whereas concurrent operations require two additional lanes shut down for construction access. Calculations were made for both screed paving (slower) and slipform paving (faster): — Using fixed forms and a screed, screed paving is usually slower, assuming the use of 7.5 yd3 agitating mixers arriving at 10 trucks/h and only one demolition crew. — Using a slipform paver, slipform paving is usually faster, assuming the use of 8.5 yd3 end dump trucks arriving at 17 trucks/h and two demolition crews. • Remove and replace with HMA. Remove the existing pavement and replace with the same depth of new HMA pavement. The roadway lanes being paved are fully shut down, only one paver with a 12-ft-wide screed is used, and HMA is paved in TABLE 1.38. MULTICOOL INPUT PARAMETERS FOR HMA OPTIONS Input Value Constant Inputs in All Scenarios Start time 1000, 7/15/2010 Environmental Conditions Ambient air temperature 60°F Average wind speed 5 mph Sky conditions Clear and dry Latitude 38° North Existing Surface Material type Granular base Moisture content Dry State of moisture Unfrozen Surface temperature 60°F Mix Specifications Mix type Dense graded PG grade 64-22 Delivery temperature 300°F Stop temperature 140°F Lift Thicknesses 3 in. of HMA total 2 lifts of 1.5 in. each 6 in. of HMA total 3 lifts of 2 in. each 9 in. of HMA total 3 lifts of 2 in., 1 lift of 3 in. 12 in. of HMA total 2 lifts of 1.5 in., 3 lifts of 3 in.

70 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE 0 2 4 6 8 10 12 14 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure (not feasible) 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure 0 2 4 6 8 10 12 14 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure (not feasible) 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure Figure 1.35. Productivity estimates for remove and replace with PCC (fixed form) using sequential operations. Solid lines indicate averages and dashed lines indicate 95% confi- dence intervals. Note: This option is not feasible using 10-h night closures. Figure 1.36. Productivity estimates for remove and replace with PCC (slipform) using sequential operations. Solid lines indicate averages and dashed lines indicate 95% confi- dence intervals. Note: This option is not feasible using 10-h night closures.

71 PROJECT ASSESSMENT MANUAL 0 2 4 6 8 10 12 14 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure (not feasible) 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure 0 2 4 6 8 10 12 14 0.0 2.0 4.0 6.0 8.0 10.0 12.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure (not feasible) 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure Figure 1.37. Productivity estimates for remove and replace with PCC (fixed form) using concurrent operations. Solid lines indicate averages and dashed lines indicate 95% confidence intervals. Note: (1) This option is not feasible using 10-h night closures, and (2) doing demolition and paving concurrently results in significantly higher productivities than doing them sequentially. Figure 1.38. Productivity estimates for remove and replace with PCC (slipform) using concurrent operations. Solid lines indicate averages and dashed lines indicate 95% confidence intervals. Note: (1) This option is not feasible using 10-h night closures, and (2) doing demolition and paving concurrently results in significantly higher productivities than doing them sequentially.

72 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE 0 2 4 6 8 10 12 14 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure 0 1 2 3 4 5 6 7 0.0 5.0 10.0 15.0 20.0 25.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure Figure 1.39. Productivity estimates for remove and replace with HMA. Solid lines indi- cate averages and dashed lines indicate 95% confidence intervals. Figure 1.40. Productivity estimates for mill and fill with HMA. Solid lines indicate aver- ages and dashed lines indicate 95% confidence intervals.

73 PROJECT ASSESSMENT MANUAL 0 1 2 3 4 5 6 7 8 9 10 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure 0 2 4 6 8 10 12 14 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure 10-hour Night Closure (not feasible) 55-hour Weekend Closure 168-hour 24/7 Continuous Week-Long Closure Figure 1.41. Productivity estimates for crack, seat, and overlay. Solid lines indicate averages and dashed lines indicate 95% confidence intervals. Figure 1.42. Productivity estimates for PCC unbonded overlay. Solid lines indicate aver- ages and dashed lines indicate 95% confidence intervals.

74 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE lifts. Lifts are generally 3 in. thick with the exception of the top two lifts, which are either 2 or 1.5 in. thick. A lift is paved for each lane across before the next lift is paved on any lane. • Mill and fill with HMA. Remove a predetermined thickness from the existing pave- ment with an HMA milling machine, then replace the same thickness with new HMA. The roadway lanes being paved are fully shut down, only one paver with a 12-ft-wide screed is used, and HMA is paved in lifts. Lifts are generally 3 in. thick with the exception of the top two lifts, which are either 2 or 1.5 in. thick. A lift is paved for each lane across before the next lift is paved on any lane. • Crack, seat, and overlay. Crack and seat the existing PCC pavement, then overlay with HMA. The roadway being paved is fully shut down, only one paver with a 12-ft-wide screed is used, and HMA is paved in lifts. Lifts are generally 3 in. thick with the exception of the top two lifts, which are either 2 or 1.5 in. thick. A lift is paved for each lane across before the next lift is paved on any lane. • Use unbonded PCC overlay. Prepare the surface of the existing PCC pavement, then overlay with PCC that is not bonded to the existing pavement. This is es- sentially like the “remove and replace with PCC” method without the demolition component. 0 2 4 6 8 10 12 14 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 Th ic kn es s of P av em en t C on st ru ct ed ( in ch es ) Lane Miles Constructed in One Closure Crack, Seat and Overlay (CSOL) Remove-and-Replace with HMA Remove-and-Replace with PCC (Fixed Form) Remove-and-Replace with PCC (Slipform) Figure 1.43. A productivity comparison of PCC remove and replace (both fixed form and slipform), HMA remove and replace, and crack, seat, and overlay (CSOL).

75 PROJECT ASSESSMENT MANUAL Analysis Tools: Traffic Impacts There are a number of analysis tools available to assist in work-zone traffic impacts es- timation. FHWA divides these tools into six broad categories (Hardy and Wunderlich, 2008a; summarized in Table 1.39): 1. Sketch-planning tools are specialized models designed for work-zone analysis. These models can vary from simple spreadsheet calculations to general delay esti- mation tools. Typically, models are deterministic and based on simple queuing equations or volume-to-capacity relationships from the HCM. Such simple estima- tion tools are often adequate for work-zone delay estimation. 2. Travel demand models are used to forecast future traffic demand based on current conditions, and to forecast future predictions of household and employment centers (Alexiadis, Jeannotte, and Chandra, 2004). Travel demand models are usu- ally used in large regional planning efforts. In work-zone analysis, they can help predict regionwide impacts of extended roadway closures (e.g., closing a freeway for several months). It is not likely that a travel demand model would be built for the specific purpose of work-zone traffic analysis. Rather, an existing model may be used if available and warranted. 3. Traffic signal optimization tools are used to develop signal timing plans. These can be useful if a temporary signal is used or if signals are retimed to accommodate work-zone traffic or increased detour route traffic. 4. Macroscopic simulation models are based on the deterministic relationships of traffic speed, flow, and density (Alexiadis, Jeannotte, and Chandra, 2004). These models treat flow as an aggregate quantity in a defined area and do not track indi- vidual vehicles. They are useful for modeling larger-area impacts of work zones because of their aggregate nature. TABLE 1.39. TRAFFIC MODEL TYPES FOR WORK-ZONE TRAFFIC IMPACTS Model Type Examples Strengths Weaknesses Sketch planning HDM, QUEWZ-98, QuickZone, CA4PRS Low cost, specific to work zones, fast Limited modeling ability, not well supported Travel demand EMME/2, TransCAD, TRANSIMS Can model large areas Low detail, cannot model short-term work-zone effects Signal optimization PASSER, Synchro Models signal timing and coordination Does not model other things Macroscopic BTS, KRONOS, METACORE/ METANET, TRANSYT-7F Can model large areas Low detail, cannot model short-term work-zone effects Mesoscopic CONTRAM, DYNASMART, DYNAMIT, MesoTS Good compromise between macro- and micromodels Data intensive Microscopic CORSIM, VISSIM, PARAMICS Can model small details, good communication tool Data intensive

76 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE 5. Mesoscopic simulation models represent the relative flow of vehicles on a network, but they do not model individual lanes or vehicles. These models are between macroscopic and microscopic models in detail and can simulate both large geo- graphic areas as well as specific corridors. They do not, however, possess the detail to model more modified strategies such as signal timing. These models require large amounts of data. 6. Microscopic simulation models simulate the movement of individual vehicles. These models require large amounts of data and can get unwieldy when a user attempts to simulate a large network. Often these models can provide animated output that can clearly communicate to decision makers and the public what the potential traffic impacts of modeled actions will be. The most appropriate modeling approach depends on the following (Hardy and Wunderlich, 2008b): • Work-zone characteristics affect the expected level of impact a work zone will have on travelers and include the geographic scale of the affected area and the complexity of the road network within this area. • Transportation management plan strategies are the means by which traffic will be managed, including such items as lane closures, full roadway closures, lane shifts, counterflow traffic, night/day work, detours, and weekend work. • Data availability and quality are the type, amount, accuracy, and timeliness of available data. • Agency resources include the owner-agency’s funding, technical staff, and schedule. • Work-zone performance measures are selected by the owner-agency to quantify traffic impacts. Typically these measures are some form of delay (in minutes or cost) either in total (total delay/cost) or peak (longest queue, longest wait). Because the use of modeling tools beyond sketch-planning tools will almost surely require traffic expertise beyond the pavement profession, further discussion is limited to a few sketch-planning tools that may be of use: QuickZone and CA4PRS. Both of these tools can provide meaningful traffic impact estimates for a relatively small mon- etary and time investment. QuickZone 2.0. A Microsoft Excel–based tool (requires Excel 97 as a minimum) that estimates work-zone traffic impacts. It allows the user to input a node-and-link network (see Figures 1.44 and 1.45) and then assign traffic counts to that network. It can coarsely simulate traffic variations between days of the week, and months of the year by applying multiples to standard ADT inputs. It can simulate multiple lane closures over time, model traffic over an entire week (Figure 1.46), and display various traffic impact metrics (Figure 1.47). These capabilities are helpful because they allow QuickZone to show differences in traffic impacts between nights and days, weekends and weekdays, and seasons (e.g., summer versus fall work). The user guide explains the algorithm QuickZone uses to estimate delay and user cost, but specific equations are not listed or discussed. QuickZone is inexpensive (about $200) but is getting relatively old (version 2.0 was released in 2005) without any significant upgrade or

77 PROJECT ASSESSMENT MANUAL Figure 1.44. A simple network that works quite well in QuickZone. Figure 1.45. A complex network simulation in QuickZone (I-5 in the Seattle, Washington, area is shown). This network simulation exposed several program bugs, was unwieldy to process, and required tedious troubleshooting to make operational. This level of complexity is not recommended.

78 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Figure 1.47. QuickZone 2.0 summary tables showing available traffic impact metrics. 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 Q ue ue L en gt h (M ile s) Time of Day After Case Queue Length (Miles) for Inbound Direction from Phase Monday 8am to Saturday midnight Reconstruction area (70-73.64) Figure 1.46. Unedited QuickZone 2.0 simulation output chart for a 1-week time period. Note that the automatic graph labeling on the horizontal axis is unreadable; however, this can be corrected by editing the graph in Excel. Period with highest delay in After Case Phase midnight to Thursday 8am Direction Inbound Day/Time Sunday 21:00 Max Queue (Miles) Max Delay (min) Passenger Cars Truck Detour Econ/Misc Total Baseline 0 0 $0 $0 $0 $0 $0 After 34.13 778.23 $22,036,199 $1,851,536 $0 $0 $23,887,735 Total 34.13 778.23 $22,036,199 $1,851,536 $0 $0 $23,887,735 Total Project User Cost ($)

79 PROJECT ASSESSMENT MANUAL support beyond a user guide. Simple scenarios with just a few links and nodes are rela- tively easy to simulate; however, more complex scenarios become cumbersome due to tedious data entry and difficult input troubleshooting if outputs are suspect. CA4PRS. A Microsoft Access–based software tool that can be used to analyze high- way pavement rehabilitation strategies including productivity, project scheduling, traffic impacts, and initial project costs based on input data and constraints supplied by the user. The traffic impacts analysis portion of CA4PRS (labeled “Work-Zone Analysis” in the software) can simulate 24 h of traffic through a defined work zone. Work zones are defined by the number of lanes closed, the closure duration, and the work-zone capacity (Figure 1.48). Traffic can be entered by hourly count or ADT can be entered and then distributed over 24 h using hourly factors. CA4PRS can simulate a one-lane-closure sce- nario over a 24-h period. Longer closures are estimated by multiplying the results of one 24-h analysis by the total number of closures. The 24-h simulation limit using only one traffic count makes it difficult to account for longer closures (e.g., over several weeks or months), where traffic flow is likely to change over time (e.g., weekday versus weekend or summer versus fall). Output is similar to that of QuickZone (Figures 1.49 and 1.50). 2014.05.09 01 R23 Guide Chapter 1-final for composition.docx 133 [Insert Figure 1.48] Figure 1.48. CA4PRS Work-Zone Analysis input screen. [Insert Figure 1.49] Figure 1.48. CA4PRS Work-Zone Analysis input screen.

80 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Figure 1.50. CA4PRS Work-Zone Analysis hourly traffic results graph showing demand versus capacity. Figure 1.49. CA4PRS Work-Zone Analysis summary results screen showing available traffic impact metrics.

81 PROJECT ASSESSMENT MANUAL Currently, the CA4PRS user manual does not explain the delay estimation algorithm it uses. As of 2010, CA4PRS development is ongoing and licenses for state DOTs are free. CA4PRS only models traffic in the work zone and does not model any wider network. LIFE-CYCLE ASSESSMENT (ENVIRONMENTAL ACCOUNTING) Purpose This section overviews a method for determining the inputs and outputs of a pave- ment system that are relevant to the environment. This can include, but is not limited to, energy use, water use, emissions, raw materials, and human health impacts. This method, called life-cycle assessment (LCA), is essentially an environmental account- ing protocol. LCA results can be used as part of the decision-making process when determining the appropriate pavement rehabilitation or reconstruction strategy. For instance, if an owner-agency must comply with a greenhouse gas (GHG) reduction mandate, options resulting in less GHG may be considered more favorably. Often, but not always, environmental accounting results tend to agree with life-cycle assessment results in pavement construction scenarios. In the future, it is likely that energy and emissions associated with roadway con- struction, or any industry, will be scrutinized more carefully. GHG emissions are likely to be subject to a cap-and-trade scheme in the United States and are increasingly being addressed through the National Environmental Policy Act (NEPA) as a recent White House Council on Environmental Quality guidance shows (Sutley, 2010). As this scru- tiny increases, there will likely be more tools to help in analysis. It also seems plausible that, once industry has a fair idea what energy, emissions, and other resources are asso- ciated with roadway construction, it will begin to adopt (either voluntarily or by regu- lation) efficiency standards associated with these items similar to what has happened with the automobile industry (i.e., fuel efficiency standards), power generation (i.e., clean energy portfolio requirements), and even toilets (i.e., maximum allowable flow). Measurement Methods An LCA attempts to identify inputs and outputs of a system that are relevant to the environment from its inception to its ultimate disposition. This means that an LCA includes everything from gathering raw materials to the point at which those mate- rials are returned to the environment. This collection of all processes from “cradle to grave” allows LCA to provide a cumulative total of inputs and outputs (e.g., energy, emissions, water, use) for a final product and the environmental impacts associated with those inputs and outputs. The resulting environmental impacts of these cumula- tive inputs and outputs is assessed, and results can be used to compare alternatives and improve the system. The International Standards Organization (ISO) outlines a systematic four-phase approach: 1. Goal and scope help define the reasons for carrying out the LCA, the intended audi- ence, geographic and temporal considerations, system functions and boundaries, impact assessment, and interpretation methods.

82 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE 2. Inventory assessment is performed to quantify life-cycle energy use, emissions, and land and water use for technology use in each life-cycle stage. 3. Impact assessment is used to estimate the impacts of inventory results. 4. Interpretation involves investigation of the contribution of each life-cycle stage and technology use throughout the life cycle, and includes data quality, sensitivity, and uncertainty analyses. LCA in general, and for pavements in particular, is still in a relatively early stage of development and thus common practices are still developing and available data can be sparse. This presents problems when using LCA as a decision-support tool, especially when comparing alternatives. Results using different data sets, methods, and prac- tices can be an order of magnitude different for the same analyzed pavement section. Common issues with LCA include the following: • Data sources. Often LCA data come from a select few databases, such as the U.S. Life-Cycle Inventory Database (from the National Renewable Energy Laboratory), ecoinvent, and the ELCD database. These are generally reviewed for accuracy or errors and can help standardize information for use in LCAs. However, data usually come from many different sources, ranging from personal observation to national databases, which can lead to problems when comparing one LCA with another. For instance, the CO2 associated with HMA production is not a universal constant but rather varies depending on plant type, components and manufacturer, aggregate moisture content, fuel type, amount of RAP included, asphalt binder grade, crude oil source, regional electricity mix, and so forth. Although databases of national averages can lead to some consistency in results between LCAs, they often do not provide the detail necessary to distinguish between process changes (e.g., using warm-mix asphalt or not). At the very least, an LCA should clearly identify its data sources. • Missing data. There are many industrial processes where some, if not all, relevant data are not known, recorded, or made available for public use. For instance, the amount of fugitive dust on site associated with pavement construction is not generally known. Or, the exact chemical makeup of an asphalt modifier may be a trade secret that the manufacturer is not willing to divulge. • Outdated data. Sometimes data exist but are outdated. Over time, processes change, equipment improves, raw material sources change, and so forth. For ex- ample, one of the more comprehensive sources for asphalt refining comes from Eurobitume and was produced in 2000. • Data specificity. Although general average data may be more readily available or may lead to more consistency between LCAs, these data often do not contain the detail needed to distinguish between two alternatives being considered. For in- stance, the Environmental Protection Agency’s AP-42 document contains average emissions data for asphalt plants; however, it assumes only an average amount of RAP being used at the plant. Therefore, if these data are used it cannot distinguish between a mix using all virgin materials and one using 25% RAP.

83 PROJECT ASSESSMENT MANUAL • Setting boundaries. An LCA that attempts to account for all processes associated with a system can quickly become intractable. For example, one could account for the slipform paver and its energy use and emissions associated with a concrete pavement. One could also account for the energy and emissions associated with manufacturing that slipform paver. However, that leads to potentially considering the energy and emissions associated with the manufacture of the machines that made the paver, and so on. Because of this, every LCA has a defined boundary that details which processes are included and which are not. Inclusions and exclusions are often not consistent between LCAs and can be controversial. For instance, in a pavement LCA one can choose whether to include the effect of material stiffness on the rolling resistance it offers to vehicles that travel upon it. Reduced rolling resistance over the life of a pavement may lead to substantial energy savings when summed over the millions of vehicles that may use the road. • Procedural practices. Most LCAs generally follow ISO 14040 and ISO 14044. However, these standards are still quite generally written and leave much room for interpretation. No set of more precise LCA procedures exists for pavements. Despite these limitations, LCA can provide meaningful results and aid the project decision process. LCA Methods There are two main methods typically used for LCA: the process-based approach and the economic input-output (EIO)-based approach. Both methods are acceptable for per- forming LCAs, although each has its strengths. Each method is briefly discussed here. In process-based LCA, a selected system is chosen and defined so that it meets a set of desired requirements (e.g., a pavement structure to meet traffic, environmental, and structural requirements). This system is then broken down into separate processes (e.g., aggregate production, cement production, concrete transport) whose energy require- ments and emissions can be quantified. Further contributory processes can be defined and analyzed (e.g., manufacture of the aggregate crushers used in aggregate produc- tion) but at some reasonable point a “boundary” must be established beyond which no downstream contributory processes are considered. The location of this boundary is an important part of an LCA because it may significantly affect the results. Ultimately, boundary locations are somewhat subjective, which can lead to difficulty in comparing one LCA’s results to another. Process-based LCAs are desirable because they can be done in enough detail so that they include processes that can differentiate between two options (e.g., using warm-mix asphalt or not). They are problematic because of the subjective boundary and difficulty in obtaining data on specific processes. Economic input-output LCA (EIO-LCA) overcomes the subjective boundary issue and data availability issue by basing processes and their relationships on a national economic input-output model. An EIO model divides the economy of a country into industry-level sectors that represent individual activities in the selected economy and depicts the economic interaction of industries (sectors) in a nation (or a region) by showing how output of each sector is used as input for another. The system bound- ary is inherently the whole country’s entire economy. Interactions are represented by

84 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE monetary value in a matrix form, called an economic input-output table (I-O table). The data stored in the table are collected by public agencies (e.g., the Department of Commerce) during a certain time period (usually 5 years). This process conveniently avoids collecting individual process data and sets a consistent boundary (the nation’s economy). EIO-LCA can be problematic because it uses aggregate data, which can be inconsistently aggregated or do not contain enough detail to differentiate between two options (e.g., using warm-mix asphalt or not). Typical Values There have been a number of documented pavement LCAs in the past decade or so that can provide valuable information on typical values. Muench (2010) reviewed 12 pavement LCA papers and reports that documented 66 assessments of actual or hypo- thetical roadways and found the following: • System scope. Most LCAs tend to address the pavement structure only and not include other road features (e.g., striping, guardrails). Analysis periods are usually 40 to 50 years. • Relation of roadway construction to traffic use. A good rule of thumb is that the energy expended in initial construction of a new roadway is roughly equivalent to the energy used by traffic on the facility over 1 to 2 years. • Relation of roadway construction to operations. Operations are defined as those equipment, actions, and operations that happen on a routine basis necessary to ensure proper and safe roadway use. They include items such as lighting, traffic signals, deicing, sanding, drawbridge actions, and toll booths. Construction energy ranges from about 25% to 100% of operations energy. • Total energy use. It can be loosely stated that energy expenditures per lane mile of pavement are typically on the order of 3 to 7 TJ depending on the pavement sec- tion, maintenance activities, and LCA scope. • CO2 emissions. It can be loosely stated that CO2 emissions per lane mile of pave- ment are typically on the order of 200 to 600 tons depending on the pavement section, maintenance activities, and LCA scope. • Contribution of roadway construction components. The following general state- ments are reasonable: — Materials production accounts for about 60% to 80% of energy use and 60% to 90% of CO2 emissions. — Construction accounts for less than 5% of energy use and CO2 emissions. — Transportation associated with construction accounts for about 10% to 30% of energy use and about 10 percent of CO2 emissions. — Maintenance activities account for a broad range of about 5% to 50% of energy and CO2 emissions.

85 PROJECT ASSESSMENT MANUAL Analysis Tools At present, there are few tools available to help the nonspecialist conduct a meaning- ful pavement LCA; however, several efforts are under way to develop such tools. This section briefly overviews the few existing tools. EIO-LCA is an online tool from Carnegie Mellon University’s Green Design Insti- tute (www.eiolca.net) that uses the EIO method to report U.S. economic sector aver- ages of economic activity, greenhouse gases, energy, toxic releases, and water use for different processes (Figure 1.51). Answers for specific sectors can be obtained quickly; however, there is not enough detail to distinguish between processes within a sector (e.g., using warm-mix asphalt or not). PaLATE is a Microsoft Excel–based tool from the University of California, Berkeley’s Consortium on Green Design and Manufacturing that allows the user to input pave- ment construction and materials parameters and calculates life-cycle energy use and a number of life-cycle emissions parameters (Figure 1.52). It is primarily built on the EIO-LCA method, but it uses the process approach for a few items. PaLATE contains numerous errors in process data, computation, and physical input parameters. These errors are significant enough to cause results to be incorrect by orders of magnitude Figure 1.51. Output screen of the EIO-LCA online tool showing greenhouse gases associated with $1 million of economic activity in sector 230230 (highway, street, bridge, and tunnel construction) using the 1997 Industry Benchmark Model for producer prices.

86 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE in some cases, thus rendering PaLATE essentially useless. Recently, the University of Washington has rebuilt and simplified PaLATE (Figures 1.53 and 1.54) for interim use in their performance metric and has posted a working version on their website (www.greenroads.us). This version has not been validated by any outside party. CHANGER is a computer software program from the International Road Fed- eration (IRF) that calculates the life-cycle CO2 emissions associated with pavement construction. It uses a process-based method and has been analyzed and validated by Figure 1.52. Introduction screen for PaLATE from the Consortium on Green Design and Manufacturing at the University of California, Berkeley.

87 PROJECT ASSESSMENT MANUAL Figure 1.53. Introduction screen for PaLATE as modified by the University of Washington for use with Greenroads. the Traffic Facilities Laboratory (LAVOC) of the Swiss Federal Institute of Technol- ogy, or Ecole Polytechnique Fédérale de Lausanne (EPFL). At present it only reports CO2 emissions. The IRF plans to expand this tool to address the entire roadway (i.e., beyond just the pavement to include signs, striping, guardrail, etc.).

88 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Example LCA Using PaLATE as Modified by the University of Washington for Greenroads A local collector road in Kailua, Hawaii, is scheduled for repaving. The work essen- tially involves removing 6 in. of HMA with a milling machine and replacing it with two layers of HMA: a 4-in. base course and a 2-in. surface course. Initial construction quantities are as follows: • Surface course: 9,516 tons of HMA. — 5.5% asphalt by total weight of mix. — No recycled material in the mix. Figure 1.54. Screenshot of PaLATE output as modified by the University of Washington for use with Greenroads.

89 PROJECT ASSESSMENT MANUAL • Base course: 18,790 tons of HMA. — 5% asphalt content by total weight of mix. — 10% glass cullet by total weight of mix. • Tack coat: 0.15 gallons/yd2 over 79,386 yd2 = about 61 yd3 of asphalt emulsion. • Milling: 79,386 yd2 of 6-in.-deep milling. An LCA is performed using a 50-year analysis period and assuming a 2-in. mill and fill every 10 years (years 10, 20, 30, and 40). Preservation mill-and-fill quantities are as follows: • Surface course: 7,913 tons of HMA. — 5.5% asphalt by total weight of mix. — No recycled material in the mix. • Tack coat: 0.15 gallons/yd2 over 79,386 yd2 = about 61 yd3 of asphalt emulsion. • Milling: 79,386 yd2 of 2-in.-deep milling. Materials for this process come from the following locations: • Aggregate, HMA, and RAP from a local quarry 6 mi from the job site. • Asphalt from a local asphalt terminal 30 mi from the job site. Results from this analysis are as follows (Figure 1.54): • 29,323.4 GJ of life-cycle energy consumption. • 4,446,563 kg of life-cycle CO2 equivalent emissions. Using this baseline scenario, several other options were investigated to determine LCA impacts: • Remove the glass cullet from the base course (No Glass Cullet). • Include 10% RAP in the surface and base courses (10% RAP). • Include 20% RAP in the surface and base courses (20% RAP). • Include 20% RAP in the surface course and 40% RAP in the base course (20% surface/40% base RAP). • Include 30% RAP in the surface course and 40% RAP in the base course (30% surface/40% base RAP). • Use warm-mix asphalt assuming a 20% reduction in energy and CO2 emissions from the HMA manufacturing process only (WMA). • Use a stone-matrix asphalt (SMA) surface course at 6.5% asphalt by total weight of mix that allows a surface life of 15 years. This results in resurfacing at years 15 and 30 only. • Use an ultimate combination of a SMA surface course, no glass in the base course, 40% RAP in the base course, and warm-mix asphalt for both courses (Ultimate).

90 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Figures 1.55 through 1.57 show the percentage change from the baseline practice in terms of energy consumption. Some general conclusions that can be reached using this example are the following: • Extending service life can be the biggest single influence in energy used and CO2 emitted by the pavement. • Often, a combination of options can produce an even greater savings in energy used and CO2 emitted by the pavement. • The inclusion or exclusion of the glass cullet (inclusion is a state requirement) makes very little difference in energy used and CO2 emitted by the pavement. 29.3 29.4 28.8 28.2 27.9 26.5 27.6 25.1 23.3 0 5 10 15 20 25 30 35 Current Practice No Glass Cullet 10% RAP in surface and base 20% RAP in surface and base 20% surface/40% base RAP 30% surface/40% base RAP WMA SMA Surface Ultimate Life Cycle Energy Consumption (TJ) Figure 1.55. Life-cycle energy consumption for the current practice and eight alternate scenarios for the example LCA.

91 PROJECT ASSESSMENT MANUAL 4446 4452 4391 4336 4304 4192 4322 4128 3968 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Current Practice No Glass Cullet 10% RAP in surface and base 20% RAP in surface and base 20% surface/40% base RAP 30% surface/40% base RAP WMA SMA Surface Ultimate Life Cycle CO2e Emissions (Tonnes) Figure 1.56. Life-cycle CO2 equivalent emissions for the current practice and eight alternate scenarios for the example LCA.

92 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Figure 1.57. The percentage change from the baseline value of energy consumption for a number of alternate scenarios for the example LCA. 0.3% -1.7% -3.8% -4.8% -9.6% -5.8% -14.3% -20.5% -25.0% -20.0% -15.0% -10.0% -5.0% 0.0% 5.0% No Glass Cullet 10% RAP in surface and base 20% RAP in surface and base 20% surface/40% base RAP 30% surface/40% base RAP WMA SMA Surface Ultimate Percentage Change in Energy Consumption from Baseline Practice Further Reading on Pavement LCAs Athena Institute. “A Life-Cycle Perspective on Concrete and Asphalt Roadways: Embodied Primary Energy and Global Warming Potential,” Report to the Cement Association of Canada, 2006. Carpenter, A. C., K. H. Gardner, J. Fopiano, C. H. Benson, and T. B. Edil. Life Cycle Based Risk Assessment of Recycled Materials in Roadway Construction. Waste Management, Vol. 27, 2007, pp. 1458–1464. Chui, C-T., T-H. Hsu, and W-F. Yang. Life Cycle Assessment on Using Recycled Materi- als for Rehabilitating Asphalt Pavements. Resources, Conservation and Recycling, Vol. 52, 2008, pp. 545–556. Horvath, A. “Life-Cycle Environmental and Economic Assessment of Using Recycled Materials for Asphalt Pavements,” University of California Transportation Center, Berkeley, Calif., 2003. Huang, Y., R. Bird, and M. Bell. A Comparative Study of the Emissions by Road Mainte- nance Works and the Disrupted Traffic Using Life Cycle Assessment and Micro-simulation. Transportation Research Part D, Vol. 14, 2009, pp. 197–204.

93 PROJECT ASSESSMENT MANUAL Huang, Y., R. Bird, and O. Heidrich. Development of a Life Cycle Assessment Tool for Construction and Maintenance of Asphalt Pavements. Journal of Cleaner Production, Vol. 17, 2009, pp. 283–296. Mroueh, U.-M., P. Eskola, and J. Laine-Ylijoki. Life-Cycle Impacts of the Use of Industrial By- Products in Road and Earth Construction. Waste Management, Vol. 21, 2001, pp. 271–277. Rajendran, S., and J. A. Gambatese. Solid Waste Generation in Asphalt and Reinforced Concrete Roadway Life Cycles. Journal of Infrastructure Systems, Vol. 13, No. 2, 2005, pp. 88–96. Stripple, H. “Life Cycle Inventory of Asphalt Pavements,” IVL Swedish Environmental Research Institute Ltd. report for the European Asphalt Pavement Association (EAPA) and Eurobitume, 2000. Stripple, H. “Life Cycle Assessment of Road: A Pilot Study for Inventory Analysis, Second Revised Edition,” IVL Swedish Environmental Research Institute Ltd. report for the Swedish National Road Administration, 2001. Tramore House Regional Design Office. “Integration of the Measurement of Energy Usage into Road Design,” Final Report to the Commission of the European Directorate-General for Energy and Transport, Project 4.1031/Z/02-091/2002, 2006. Treloar, G. J., P. E. D. Love, and R. H. Crawford. Hybrid Life-Cycle Inventory for Road Construction and Use. Journal of Construction Engineering and Management, Vol. 130, No. 1, 2004, pp. 43–49. Weiland, C. D. “Life Cycle Assessment of Portland Cement Concrete Interstate Highway Rehabilitation and Replacement,” master’s thesis, University of Washington, Seattle, Wash., 2008. Zapata, P., and J. A. Gambatese. Energy Consumption of Asphalt and Reinforced Concrete Pavement Materials and Construction. Journal of Infrastructure Systems, Vol. 11, No. 1, 2005, pp. 9–20. MISCELLANEOUS MATERIAL PROPERTIES Purpose This section provides summaries of material properties that are relevant in designing pavement renewal options. Material Properties Table 1.40 shows typical layer moduli for several material conditions. Table 1.41 shows information about rubblized PCC and Table 1.42 about crack-and-seat renewal.

94 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE REFERENCES AASHTO. “AASHTO Guide for Design of Pavement Structures,” American Association of State Highway and Transportation Officials, Washington, D.C., 1993. AASHTO. “Standard Practice for Determining Maximum Rut Depth in Asphalt Pavements,” AASHTO Designation PP38-00, 2000. AASHTO. “Standard Practice for Quantifying Cracks in Asphalt Pavement Surfaces,” AASHTO Designation PP44-01, 2001. Alexiadis, V., K. Jeannotte, and A. Chandra. “Traffic Analysis Toolbox Volume I: Traffic Analysis Tools Primer,” FHWA-HRT-04-038, FHWA, U.S. Department of Transportation, 2004. Borchardt, D. W., P. Geza, D. Sun, and L. Ding. “Capacity and Road User Cost Analysis of Selected Freeway Work Zones in Texas,” FHWA/TX-09/0-5619-1, Texas Department of Transportation, Austin, 2009. http://tti.tamu.edu/documents/0-5619-1.pdf. Buncher, M., G. Fitts, G.,T. Scullion, T., and R. McQueen, R. (2008), “Development of Guidelines for Rubblization—Final Report,” Airfield Asphalt Pavement Technology Pro- gram Project 04-01, Asphalt Institute, May 2008. TABLE 1.40. HMA PAVEMENT TYPICAL MODULI AND RANGES OF MODULI Material Modulus Range (psi) HMA (temperature dependent) 50,000–4,000,000 Cracked HMA range 50,000–500,000 Cracked HMA (10% of wheelpath—slight to moderate fatigue cracks) 100,000–250,000 Pulverized HMA 40,000 TABLE 1.41. PCC PAVEMENT RUBBLIZATION TYPICAL MODULI AND RANGES OF MODULI Material Value or Property Ratio of rubblized PCC elastic modulus to original PCC slab elastic modulus 0.05 Slab modulus range before rubblization Range: 3,000,000–7,000,000 psi Typical slab modulus 4,000,000 psi Rubblized PCC modulus Range: 40,000–700,000 psi Typical rubblized PCC modulus 150,000 psi TABLE 1.42. CRACK-AND-SEAT AND BREAK-AND-SEAT RENEWAL Material Value or Property Typical modulus of crack-and-seated PCC pavement 200,000 psi Modulus of crack-and-seated PCC pavement Range: 200,000–800,000 psi Modulus of break-and-seated PCC pavement Range: 250,000–2,000,000 psi

95 PROJECT ASSESSMENT MANUAL Burnham, T. “Application of Dynamic Cone Penetrometer to Minnesota Department of Transportation Pavement Assessment Procedures,” Report MN/RC 97/19, Minnesota Depart ment of Transportation, St. Paul, 1997. Chou, Y. J., J. Uzan, and R. L. Lytton. “Backcalculation of Layer Moduli from Non- destructive Pavement Deflection Data Using the Expert System Approach,” Nondestructive Testing of Pavements and Backcalculation of Moduli, ASTM STP 1026, American Society for Testing and Materials, Philadelphia, Pa., 1989, pp. 341–354. Darter, M. I., R. P. Elliott, and K. T. Hall. “Revision of AASHTO Pavement Overlay Design Procedure,” Project 20-7/39, National Cooperative Highway Research Program, Transpor- tation Research Board, Washington, D.C., 1991. Federal Highway Administration (FHWA). “1999 Status of the Nation’s Highways, Bridges, and Transit: Conditions and Performance,” Report FHWA-PL-99-017, U.S. Department of Transportation, 1999. Federal Highway Administration (FHWA). “Full Road Closure for Work Zone Operations: A Cross-Cutting Study,” U.S. Department of Transportation, 2003. http://ops.fhwa.dot.gov/ wz/resources/publications/FullClosure/CrossCutting/its.htm. Federal Highway Administration (FHWA). “Intelligent Transportation Systems in Work Zones: Keeping Traffic Moving During Reconstruction of the Big I, a Major Interstate- Interstate Interchange in Albuquerque,” U.S. Department of Transportation, 2004. Federal Highway Administration (FHWA). “2006 Status of the Nation’s Highways, Bridges, and Transit: Conditions and Performance,” U.S. Department of Transportation, 2006. www.fhwa.dot.gov/policy/2006cpr/chap3.htm. Fwa, T. “The Handbook of Highway Engineering,” Taylor & Francis, Boca Raton, Fla., 2006. Geophysical Survey Systems, Inc. “GSSI Handbook for Radar Inspection of Concrete,” Salem, N.H., August 2006. www.geophysical.com. Hardy, M., and K. Wunderlich. “Traffic Analysis Toolbox Volume VIII: Work Zone Model- ing and Simulation—A Guide for Decision Makers,” FHWA-HOP-08-029, FHWA, U.S. Department of Transportation, 2008a. Hardy, M., and K. Wunderlich. “Traffic Analysis Toolbox Volume IX: Work Zone Model- ing and Simulation—A Guide for Analysts,” FHWA-HOP-09-001, FHWA, U.S. Department of Transportation, 2008b. Heckel, L. “Rubblizing with Bituminous Concrete Overlay—10 Years’ Experience in Illinois,” Report IL-PRR-137, Illinois Department of Transportation, 2002. Illinois DOT. “Subgrade Stability Manual,” Policy Mat-10, Illinois Department of Trans- portation, 1982. Illinois DOT. “Dynamic Cone Penetrometer,” Pavement Technology Advisory PTA-T4, Bureau of Materials and Physical Research, Illinois Department of Transportation, 2005. Kim, T., D. J. Lovell, and J. Paracha. A New Methodology to Estimate Capacity for Freeway Work Zones. Presented at the 2001 Transportation Research Board Annual Meeting, Washington D.C., 2001. http://www.workzonesafety.org/files/documents/ database_ documents/00675.pdf. Lee, E. B. “CA4PRS User Manual (Version 2.1),” University of California, Berkeley, 2008.

96 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Lee, E. B., J. T. Harvey, and D. Thomas. Integrated Design/Construction/Operations Analy- sis for Fast-Track Urban Freeway Reconstruction. Journal of Construction Engineering and Management, Vol. 131, No. 12, 2005, pp. 1283–1291. Lee, E. B., and C. W. Ibbs. Computer Simulation Model: Construction Analysis for Pave- ment Rehabilitation Strategies. Journal of Construction Engineering and Management, Vol. 131, No. 4, 2005, pp. 449-458. Lee, E. B., C. W. Ibbs, and D. Thomas. Minimizing Total Cost for Urban Freeway Recon- struction with Integrated Construction/Traffic Analysis. Journal of Infrastructure Systems, Vol. 11, No. 4, 2005, pp. 250–257. Lee, E. B., H. Lee, and J. T. Harvey. Fast-Track Urban Freeway Rehabilitation with 55 h Weekend Closures: I-710 Long Beach Case Study. Journal of Construction Engineering and Management, Vol. 132, No. 5, 2006, pp. 465–472. Lee, E. B., H. Lee, and C. W. Ibbs. Productivity Aspects of Urban Freeway Rehabilitation with Accelerated Construction. Journal of Construction Engineering and Management, Vol. 133, No. 10, 2007, pp. 798–806. Lee, E. B., J. R. Roesler, J. T. Harvey, and C. W. Ibbs. “Case Study of Urban Concrete Pave- ment Reconstruction and Traffic Management for the I-10 (Pomona, CA) Project,” Univer- sity of California, Berkeley Institute of Transportation Studies, Pavement Research Center, Report for the Innovative Pavement Research Foundation, Falls Church, Va., 2001. Lee, E. B., J. R. Roesler, J. T. Harvey, and C. W. Ibbs. Case Study of Urban Concrete Pavement Reconstruction on I-10. Journal of Construction Engineering and Management, Vol. 128, No. 1, 2002, pp. 49–56. Maze, T. H., S. D. Schrock, and A. Kamyab. Capacity of Freeway Work Zone Lane Clo- sures, Mid-Continent Transportation Symposium 2000 Proceedings, Iowa State University, Ames, May 15, 2000. http://www.ctre.iastate.edu/pubs/midcon/Maze.pdf. McCullough, B. F. “Distress Mechanisms-General,” Special Report No. 126, Highway Research Board, National Academy of Sciences, Washington, D.C., 1971. McGhee, K. “Automated Pavement Distress Collection Techniques,” Synthesis 334, National Cooperative Highway Research Program, Transportation Research Board, National Research Council, Washington, D.C., 2004. Miller, J. S., and W. Y. Bellinger. “Distress Identification Manual for the Long-Term Pave- ment Performance Program (Fourth Edition),” Report FHWA-RD-03-031, Office of Infra- structure Research and Development, Federal Highway Administration, McLean, Va., 2003. Minnesota DOT. “User Guide to the Dynamic Cone Penetrometer,” Office of Minnesota Road Research, Minnesota Department of Transportation, 1993. Muench, S. T. Roadway Construction Sustainability Impacts: A Life Cycle Assessment Review. Transportation Research Record: Journal of the Transportation Research Board, No. 2151, Transportation Research Board of the National Academies, Washington, D.C., 2010. Muench, S. T., B. Ozolin, J. Uhlmeyer, and L. M. Pierce. Rapid Concrete Panel Replace- ment in Washington State: Lessons Learned. Presented at the International Conference on Optimizing Paving Concrete Mixtures and Accelerated Concrete Pavement Construction and Rehabilitation, Atlanta, Ga., November 7–9, 2007. http://offcampus.lib.washington. edu/login?url=http://ascelibrary.aip.org/getpdf/servlet/GetPDFServlet?filetype=pdf&id= JCEMD4000128000001000049000001&idtype=cvips.

97 PROJECT ASSESSMENT MANUAL National Asphalt Pavement Association (NAPA).“Balancing Production Rates in Hot Mix Asphalt Operations,” Report IS-120, NAPA, Lanham, Md., 1996. National Highway Institute. Course 131063, “Hot Mix Asphalt Pavement Evaluation and Rehabilitation—Reference Manual,” Publication FHWA-NHI-02-002, 2003. “NCHRP Synthesis of Highway Practice 26: Roadway Design in Seasonal Frost Areas,” Transportation Research Board, National Research Council, Washington, D.C., 1974. Sebesta, S., and T. Scullion. “Field Evaluations and Guidelines for Rubblization in Texas,” Report FHWA/TX-08/0-4687-2, Texas Transportation Institute, FHWA, Washington, D.C., 2007. Shafizadeh, K., and F. Mannering. Acceptability of Pavement Roughness on Urban High- ways by Driving Public. Transportation Research Record: Journal of the Transportation Research Board, No. 1860, Transportation Research Board of the National Academies, Washington, D.C., 2003. Shahin, M. “PAVER Distress Manual,” TR 97/104, U.S. Army Construction Engineering Research Laboratories, Champaign, Ill., 1997. SHRP. “Distress Identification Manual for the Long-Term Pavement Performance Project,” National Research Council, Washington, D.C., 1993. Stark, D. “Handbook for the Identification of Alkali-Silica Reactivity in Highway Struc- tures,” Report SHRP-C-315, Strategic Highway Research Program, National Research Council, Washington, D.C., 1994. Start, M., K. Jeong, and W. Berg. Potential Safety Cost-Effectiveness of Treating Rutted Pavements. Transportation Research Record 1629, TRB, National Research Council, Wash- ington, D.C., 1998. Stubstad, R. N., Y. J. Jiang, M. I. Clevenson, and E. O. Lukanen. “Review of the Long- Term Performance Backcalculation Results—Final Report,” Report FHWA-HRT-05-150, FHWA, U.S. Department of Transportation, 2006. Sutley, N. H., Chair, Council on Environmental Quality. “Memorandum for Heads of Fed- eral Departments and Agencies: Draft NEPA Guidance on Consideration of the Effects of Climate Change and Greenhouse Gas Emissions,” February 18, 2010. Terzaghi, K., and R. Peck. Soil Mechanics in Engineering Practice, John Wiley and Sons, New York, 1967. Texas DOT. Hydroplaning, “Hydraulic Design Manual,” Texas DOT, 2009. Transportation Research Board. “Highway Capacity Manual (HCM2000),” TRB, National Research Council, Washington, D.C., 2000. U.S. Army. “Design of Aggregate Surfaced Roads and Airfields,” Technical Manual TM 5-822-12, Department of the Army, Washington, D.C., 1990. Von Quintus, H. L., and A. L. Simpson. “Backcalculation of Layer Parameters for LTPP Test Sections, Volume II: Layered Elastic Analysis for Flexible and Rigid Sections,” Report FHWA-RD-01-113, FHWA, U.S. Department of Transportation, 2002. Walls, J., and M. R. Smith. “Life-Cycle Cost Analysis in Pavement Design,” Report FHWA-SA-98-079. FHWA, 1998. http://isddc.dot.gov/OLPFiles/FHWA/013017.pdf.

98 GUIDE TO USING EXISTING PAVEMENT IN PLACE AND ACHIEVING LONG LIFE Webster, S., R. Grau, and W. Thomas. “Description and Application of Dual Mass Dynamic Cone Penetrometer,” Instruction Report GL-92-3, Waterways Experiment Station, U.S. Army Corps of Engineers, 1992. Yau, A., and H. L. Von Quintus. “Study of LTPP Laboratory Resilient Modulus Test Data and Response Characteristics,” Report FHWA-RD-02-051, FHWA, U.S. Department of Transportation, 2002.