Recycling and Reclamation of Asphalt Pavements Using In-Place Methods (2011)

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.

6 CHAPTER TWO DEVELOPMENT OF A RECYCLING PROJECT Although each of the in-place recycling processes differs in purpose, the development of a recycling project has a number of common considerations. Figure 4 outlines the steps needed for project selection, material selections, mix designs, assessment of structural capacity, and construction sequences. The following sections are organized in order of the steps outlined in this figure. Each section identifies specific points in the development processes where different considerations are needed to select the best in-place recy- cling process for a given project. FIGURE 4 Steps in selecting, designing, constructing, and specifying in-place recycling projects. PROJECT SELECTION CRITERIA Project selection is the first step in development of a recycling project and consists of an assessment of existing pavement condition, traffic, geometric and environmental consider- ations, and identification of surface treatments needed for weather (e.g., snowplows, wet roads), restriction of water penetration, traffic, and anticipated capacity improvements. The contractor members of ARRA were asked to answer the same survey questions as the state agency materials engineers. A comparison of the responses for the two pop- ulations of respondents provides insight into topics where there is good agreement and those in need of further educa- tion, research, and clarification (see Figure 5). FIGURE 5 Comparison of agency and contractor responses for measurements of pavement condition used for in-place recycling projects. Pavement Condition Functional condition of an existing pavement describes roadway features that meet the users’ need for ride quality (smoothness), safety (polishing, bleeding/flushing, friction), and geometry (e.g., lane widths). Distress measurements that influence the ride quality and safety include potholes, bumps, depressions, shoving, and slipping. Structural condition is the ability of the roadway to carry the traffic loads. Structural issues can be identified with nondestruc- tive testing such as falling weight deflectometer (FWD) and by quantifying load- and support-related distresses such as longitudinal cracking in the wheel paths, edge cracking, and fatigue cracking. Assessing Existing Pavement Condition The two most common methods of assessing the condition of the pavement are distress surveys and smoothness. Most states, independent of in-place recycling processes, use dis- tress surveys as the primary source of information for ini- tially identifying potential preservation, maintenance, and rehabilitation activities (Table 5). A more limited number of

7 states also include ride quality measurements before placing a recycling project. Specific layer properties and confirma- tion of layer thicknesses and properties are discussed in the sections on material selection and mix designs of this report. TABLE 5 AGENCY RESPONSES FOR PRECONSTRUCTION FIELD TESTING FOR IN-PLACE RECYCLING PROJECTS Preconstruction Field Testing: Before construction, I typically use: Preconstruction Work States HIR CIR FDR Condition Distress Survey AR, AZ, CA, CO, FL, ID, IA, KS, KY, MD, MO, MT, NC, NE, NY, TX, VT, WA AZ, CA, CO, CT, DE, ID, IA, KS, MD, MN, MO, MT, NE, NH, NV, NY, OR, RI, SD, UT, VA, VT, WA, WI, WV AR, AL, CA, CO, CT, DE, GA, ID, IA, MD, MN, MO, MT, NE, NH, NV, OR, SC, SD, TX, UT, VA, VT, WI, WY Ride Quality (smoothness measurements) AR, AZ, CA, CO, FL, GA, ID, KS, MD, MT, VT, WA AZ, CA, CO, ID, MD, MN, MT, NH, NV, UT, VA, VT, WA, WY AL, CA, CO, MD, MN, NH, NV, UT, VA, VT Both agencies and contractors rely on distress survey information when considering projects for recycling. These assessments are increasingly important for CIR and FDR projects. Contractors are less likely to consider ride quality when evaluating preconstruction test results. In summary, pavement condition, particularly distress surveys, is one of the most important factors in the selection of an in-place recycling method. Milling Depths The existing pavement condition and the type, extent, and severity of the distresses will indicate the depth of recycling needed for preservation, maintenance, and rehabilitation, and hence help identify the most useful recycling process. ARRA (2001) provides recommendations for the various distresses that can be addressed with a particular recycling process, along with the appropriate milling depths (Table 6). HIR projects are recommended for milling only the top 1 to 2 in., CIR from 2 to 4 in., and FDR for greater than 6 in. Table 6 includes typical distresses that can be addressed at each recycling depth. The milling depths used by state agencies on HIR proj- ects can go as deep as 4 in. when two passes are used (Table 7). Seven states mill CIR projects from 1 to 6 in. As with the other recycling processes, a limited number of agencies applied FDR outside the recommended depths. Written com- ments indicated that a maximum depth for FDR may be use- ful, as some states reported difficulty in achieving adequate compaction in lifts thicker than 12 to 14 in. Four states use TABLE 6 GENERAL GUIDELINES USES FOR IN-PLACE RECYCLING BASED ON PAVEMENT DISTRESSES PRESENT IN THE EXISTING PAVEMENT (based on ARRA 2001) Distress HIR CIR FDRSurface Recycling Remixing Repaving Milling Depths 25 mm (1 in.) X — — — — 25 to 50 mm (1 to 2 in.) X X X X — 25 to 75 mm (1 to 3 in.) — X X X — 50 to 100 mm (2 to 4 in.) — — — X X 100 to 150 mm (4 to 6 in.) — — — — X >150 mm (>6 in.) — — — — X Distresses Alligator Cracking P F G G G Bleeding, Flushing F F F F G Block Cracking F F G G G Bumps F F F F G Edge Cracking P F F F G Friction Improvement P F G G G Longitudinal Cracks (non-wheel path) F F G G G Longitudinal Cracks (wheel path) F F G G G Oxidation G G G G G Patches F G F G G Polishing P G G G G Potholes F G G G G Raveling G G G G G Rutting F G F G G Reflective Cracking F F G G G Shrinkage Cracking — — — — — Shoulder Dropoff P P P P P Shoving F G F G G Slippage F F G G G Transverse Cracks F F F G G Moisture Damage P F G G G Ride Quality (distress related) F F F F G Minor Profile Corrections F F F F G G = good process for addressing distress. F = fair process for addressing distress. P = not likely to fully address distress. This table is included as a reference for general guidelines and should not be used exclusively to select a recycling process.

8 FDR at shallower depths of 4 to 6 in., which is likely a func- tion of thin hot mix asphalt (HMA) layers common on low traffic volume roadways. Shallow depths of 2 to 4 in. reflect thin HMA layers or multiple surface treatments placed on the subgrade, which can be found on very low traffic volume roadways. TABLE 7 AGENCY RESPONSES FOR MILL DEPTHS USED ON RECYCLING PROJECTS Typical Milling Depth: Indicate the most common depth of milling for your recycling projects Mill Depths Agency Responses HIR CIR FDR 25 to 50 mm (1 to 2 in.) AR, CA, CO, FL, ID, IA, KS, KY, MD, MO, NC, NE, TX, WA NC — 50 to 100 mm (2 to 4 in.) AZ, MD, MT AZ, CA, CT, DE, ID, IL, IA, KS, MN, MO, MT, ND, NE, NV, NY, OR, SD, UT, VA, VT, WA, WY MN, NC 100 to 150 mm (4 to 6 in.) — CO, DE, IL, MO, RI, VA, WI AL, DE, MO, VT >150 mm (>6 in) — — AL, CA, CO, GA, ID, IL, IA, MT, ND, NE, NV, NY, OR, SC, TX, UT, VA, VT, WY Contractors frequently use milling depths of 50 to 100 mm (2 to 4 in.; 4 in. requires two passes) for CIR processes, with a significantly higher percentage of contractors than agencies using milling depths outside of this range (Figure 6). There is good agreement between agencies and contrac- tors on milling depths greater than 6 in. for FDR projects. A higher percentage of contractors use shallow (50 to 100 mm or 2 to 4 in.) milling depths for FDR projects than agencies. This may represent more nonstate work on low traffic vol- ume roadways by contractors. A number of state agencies and contractors use the ARRA-recommended range of recycling depths for each process; however, the actual depth of recycling can vary depending on project needs. Guidance on the maximum FDR recycling depth (i.e., lift thickness) is needed so that the desired layer density can be obtained. Agencies appear to underuse FDR for thinner layers. FIGURE 6 Comparison of milling depths used by agencies and contractors for each in-place recycling process. Traffic Traffic levels can limit the use of some recycling processes. When asphalt emulsions are used in CIR and FDR projects, the emulsion needs time to break (set) and the water needs time to evaporate before placing the surface course. During this curing time, the pavement needs to perform under traffic from 7 to 30 days. An appropriate selection of materials and additives can be used to minimize the time delay between recycling and placement of the surface course. Another con- sideration related to traffic level is the ability of the subgrade to support the weight of the presurface treatment traffic and recycling equipment. All recycling processes have been used at traffic levels up to 30,000 annual average daily traffic (AADT; Table 8); however, some states may limit the traffic for specific processes to less than 5,000 AADT. At over 30,000 AADT, agencies consider using only HIR or FDR processes. TABLE 8 TRAFFIC LEVELS FOR IN-PLACE RECYCLING PROJECTS Question: I would consider recycling a roadway with annual average daily traffic (AADT) levels of up to: AADT Type of In-Place Recycling Used HIR CIR FDR <5,000 CA, DC, FL, NE, VT, WY CT, IA, KY, NC, ND, NE, NY, VT, WY CT, IA, KY, MD, NC, NE, VT, WY 5,000 to 30,000 AR, AZ, IA, KY, MO, WA AZ, CA, CO, DE, ID, MD, MN, MO, NH, OR, RI, SD, UT, VA, WA AK, CO, DE, MN, MO, ND, NH, OR, SC, SD, UT >30,000 CO, ID, KS, MD, MT, NC, TX — CA, GA, ID, MT, NV, TX, VA, WI

9 Significant differences between agencies and contractors were seen at traffic levels of less than 5,000 and greater than 30,000 AADT. Contractors are less likely than the agencies to consider HIR and CIR for the low traffic levels (Figure 7). This may be related to the lack of adequate support for the recycling equipment on the thinner low-volume roadways. Contractors are more likely to consider any of the processes acceptable for the higher traffic levels. FIGURE 7 Influence of traffic levels on the selection of in- place recycling process. The average of the percentage of positive responses from both the agencies and contractors was used to rank accept- able geometric features of roadways for each of the in-place recycling processes (Table 9). Four categories of acceptabil- ity of each of the factors are defined: TABLE 9 MAXIMUM TRAFFIC LEVELS CURRENTLY USED FOR IN-PLACE RECYCLING METHODS AADT HIR CIR FDR <5,000 F F G 5,000 to 30,000 G G G >30,000 G G G P = Poor, lower than 10% average of agency and contractor. F = Fair, between 10% and 25% average of agency and contractor. G = Good, between 25% and 50% average of agency and contractor. VG = Very good, greater than 50% average of agency and contractor. • VG = very good and represents that more than 50% of agencies and contractors consider the factor acceptable. • G = good and represents that between 25% and 50% of agencies and contractors consider the factor acceptable. • F = fair and represents that between 10% and 25% of agencies and contractors consider the factor acceptable. • P = poor and represents that less than 10% of agencies and contractors consider the factor acceptable. The percentages were calculated using the number of agencies or contractors indicating experience with a par- ticular recycling process. The category groupings were arbi- trarily selected after reviewing general trends in responses. HIR, CIR, and FDR on roadways with AADT greater than 30,000 may be underused by agencies and overused on facilities with AADTs less than 5,000. Subgrade support for equipment needs to be considered. The reasons for the dif- ferences in acceptable traffic levels need to be explored. Roadway Geometry and Features Roadway geometry and features may also limit the use of in- place recycling processes. Different features will have vary- ing impacts depending on the recycling process. Geometry and features evaluated in this survey include • Tight turns < 12 m (40 ft) or switchbacks, • Mountainous terrains with grades exceeding 8%, • Manholes or other castings in the pavement layer, • Minor roadway-widening needs, • Superelevation or cross-slope correction required (minor profile corrections), and • Curbs and gutters. Features that limit state use of the HIR processes include tight turns, steep grades, castings, and the need for lane wid- ening. Agencies consider HIR projects needing minor pro- file (typically less than ½ in. crossfall) corrections or with curbs and gutters acceptable features (Table 10). CIR use is limited by the presence of tight turns, steep grades, and castings. CIR is not limited by needs for roadway widen- ing, limited profile corrections, and the presence of curbs and gutters. These features seem to have the least impact on selecting FDR for projects. Tight turns, mountainous ter- rains, and minor widening limit the state’s use of HIR, CIR, and to some extent FDR. Minor profile correction limits the use of HIR but is considered acceptable for both CIR and FDR projects. Curbs and gutters can be addressed with any of the recycling processes. Contractors differ in choices of acceptable geometry and features in several cases (Figures 8 and 9). Steep grades and castings present less of a concern for contractors than for agencies when using HIR processes. Contractors are less likely than agencies to consider tight turns and steep grades as acceptable features for CIR proj- ects. The majority of contractors with experience placing FDR projects feel comfortable using this process with any of the features listed in this survey. There is better agreement between contractors and agencies on the impact of the need for lane widening, minor profile corrections, and curbs and gutters (Figure 9).

10 FIGURE 9 Influence of roadway geometry and features on the selection of in-place recycling process (tight turns, steep grades, and castings). Percentages are based on the number of agencies and contractors with experience using the specific recycling process. Information found in the literature revealed mixed opin- ions about acceptable and unacceptable roadway features. In 2002, Lee et al. found the following not to be appropriate features for CIR recycling projects: • Numerous manholes or drainage outlets, • Excessively steep grades [5% and 706 m (2,316 ft)], • Heavily shaded areas, which increase cure times, • HMA layers less than 50 mm (2 in.) thick, and • Numerous branch roadway accesses (e.g., driveways). Lane and Kazmierowski (2005a) noted that one of the advantages of CIR was the ability to use this process on proj- ects with numerous entrances, side roads, and intersections because CIR would not result in raising the grade. TABLE 10 INFLUENCE OF ROADWAY GEOMETRIC AND FEATURES ON THE SELECTION OF IN-PLACE RECYCLING PROCESSES Question: I would consider recycling a roadway with: ADT Type of In-Place Recycling Used HIR CIR FDR Tight Turns (radius 12 m (<40 ft) or switchbacks DC, KS, KY DE, KY, MT, NC, NV, SD, UT AK, DE, GA, ID, KY, MO, MT, NW, NH, NV, SD, UT Mountainous Terrains with Grades Exceeding 8% KY, MT, VT DE, KY, MT, NV, UT, VT, WA AK, DE, ID, KY, MT, NC, NH, NV, UT Manholes or Other Castings Within Pavement Layer CA, FL, IL, MO CA, CT, DE, IL, ND, NH, NV, WI AK, CA, CT, DE, ID, NC, NH, NV, SC, WI Minor Roadway Widening Needs CO, FL, ID CO, DE, ID, IL, IA, KS, MO, NV, SD, UT VT, WI, WY AK, AL, CA, CO, DE, ID, IL, IA, MO, MT, NC, ND, NV, OR, SC, DS, UT, VA, VT, WI, WY Superelevation or Cross-slope Correction Required AZ, CO, FL, IA, KS, KY AZ, CO, DE, ID, IA, KS, KY, MO, MT, NE, NV, NY, RI, UT, VA, WA, WI AK, AL, CA, CO, CT, DE, GA, ID, IA, KY, MO, MT, NC, ND, NE, NV, SC, SD, UT, VA, VT, WI, WY Curb and Gutter AZ, CO, CT, FL, GA, ID, IL, KS, KY, MD, MO, MT, NC, NE, VT, WA CO, CT, DE, ID, IL, IA, KS, KY, MS, MO, MT, ND, NV, UT, VA, WA, WI, WY AK, AL, CA, CO, CT, DE, ID, IL, KY, MO, NV, SC, SD, UT, VA, WI, WY Note: Agencies could respond to all that apply. FIGURE 8 Influence of traffic levels on the selection of in- place recycling process (widening, minor profile correction, and curb and gutters). Percentages are based on the number of agencies and contractors with experience using the specific recycling process. A written response from one agency noted experi- ence with rutting problems when CIR was used on grades greater than 4%; however, the agency indicated that its experience was from 11 years ago. A single contractor noted that, depending on the extent of cross-slope correc- tion required, off-site material may be needed, which must be considered in the contract documents. This contractor noted that it considers mountainous terrains on an indi- vidual project basis. The agency and contractor responses were used to rank and summarize the responses for judging the ability of recy- cling processes to accommodate various geometry and road- way features (Table 11).

11 Fewer states use HIR and CIR processes in cold, wet cli- mates (Table 12). Agencies prefer to use HIR in hot climates, either wet or dry. FDR use is somewhat independent of cli- matic conditions. Several possible reasons were identified that limit HIR and CIR in wet weather conditions: • Rainy weather interrupts construction work, which requires moving large, slow equipment units on and off the project. Parking large equipment during con- struction is an issue because of the size and very slow speed of the equipment. • Damp or wet pavements slow the hot recycling con- struction process. • Possible performance issues exist if rainy weather sets in before work is complete. • Wet, cool weather delays the use of emulsion-based CIR and lengthens the curing time. Contractors are more likely than agencies to limit con- struction of HIR and CIR in cold, wet and to some extent hot, wet conditions (Figure 10). Contractors are more likely to consider HIR and CIR processes appropriate choices in dry, cold or dry, hot climates. A significantly higher percentage of contractors construct FDR projects in any of the climatic regions. The average agency and contractor responses are ranked to indicate the potential impact of climate on selecting an appropriate recycling process (Table 13). It should be noted that a good choice of materials used with a given recycling process can overcome some climate limitations. The useful- ness of a recycling process ultimately should be considered on a project-by-project basis. Communication between the agency and the contractor is needed to select the best options for a given climate. TABLE 11 INFLUENCE OF GEOMETRIC FEATURES ON PROJECT SELECTION Geometric Features Ranking of Acceptable Features for Recycling Projects HIR CIR FDR Tight Turns P F VG Steep Grades G G VG Castings G VG VG Widening F G VG Minor Profile Corrections G G VG Curbs and Gutters G G VG P = Poor, lower than 10% average of agency and contractor with experience. F = Fair, between 10% and 25% average of agency and contractor with experience. G = Good, between 25% and 50% average of agency and contractor with experience. VG = Very good, greater than 50% average of agency and contractor with experience. Roadway geometry and features need to be considered when selecting the most appropriate in-place recycling method(s) for a project. Further research is needed to iden- tify the reasons for the differences between agency and con- tractor responses. Climate A number of specifications contain weather restrictions on when recycling projects can be constructed. The survey explored preferences for recycling processes used in four general climate regions: • Cold and wet, • Cold and dry, • Hot and wet, and • Hot and dry. TABLE 12 AGENCY CLIMATE PREFERENCES FOR RECYCLING METHODS Question: Environmental Conditions: I would consider recycling on roadways in the following climate regions: Climate Climate conditions HIR CIR FDR Cold and Wet CA, GA, ID, IL, IA, KS, KY, MO, NC, NY, TX, VT, WA CA, DE, ID, IL, KS, KY, MN, MO, MT, NC, NH, NV, NY, RI, WI AK, CA, CO, DE, GA, ID, IL, KY, MN, MO, MT, ND, NH, NV, OR, SD, TX, WI, WY Hot and Wet AR, CA, CO, FL, GA, ID, IL, IA, KS, KY, MT, NC, NU, TX, VT, WA CA, CO, DE, ID, IL, IA, KS, KY, MT, NC, NH, NB, NY, RI, VA, VT, WA CO, DE, GA, ID, IL IA, KY, MD, MT, ND, NH, NV, OR, SC, SD, TX, VA, WY Cold and Dry AZ, CA, CO, GA, ID, IL, IA, KS, NY, TX, VT, WA AS, CA, CO, DE, ID, IL, KS, MT, NH, NV, NY, RI, SD, UT, VT, WA, WY CA, CO, DE, GA, ID, IL, IA, MK, MT, ND, NH, NV, OR, SD, TX, UT, VT, WY Hot and Dry AR, AZ, CA, CO, DC, FL, GA, ID, IL, IA, KS, MS, MT, NE, NY, TX, VT, WA AL, AR, AZ, CA, DE, FL, IL, IN, KS, MD, MO, NE, NJ, OR, SD, TX, UT, VA, VT, WA, WI, WY AL, AR, AZ, CA, CL, DE, FL, IL, MO, ND, NE, NJ, OR, RI, SD, TX, UT, VA, VT, WA, WI

12 Surface Treatment Selection Surface treatments for HIR and CIR projects are commonly selected on the basis of climate considerations (e.g., drainage of surface water, providing impermeable membranes, snow- plows), noise reduction, and friction improvement. In some cases, overlays are selected to improve the structural capacity of the roadway. Survey results show that a structural overlay is commonly used as the final wearing surface for all types of recycling projects (Table 14). If an overlay is not used, then a combination fog and chip seal is the next most popular sur- face. Several states reported using slurry seal or microsurface treatments, but none use a fog seal by itself, a microsurface, or an open-graded friction course (OGFC) for HIR projects. The written responses indicate that other surfacings have been used, including a chip seal (without fog), rubberized OGFC, stone matrix asphalt, and unreinforced concrete overlay. An additional question was included in the surveys to assess the impact of climate conditions on the selection of the surface treatment. Only a few states indicated that cli- mate is a consideration. Agencies rely more on the use of traffic [AADT, equivalent single-axel loads (ESALs), per- centage of trucks], functional classification of roadway, existing distresses, expected performance of surface course mix, cost, and experience. Surfaces placed by contractors for FDR projects are fre- quently a structural overlay, integral overlay, or a fog–chip combination. However, contractors use a wider range of other surface treatments (Figure 11) for all types of recycling projects. A key criterion for two contractors was whether snowplows would be used on the surface mix. Contractors consider traffic, structural designs, existing distresses, per- formance (raveling, texture, ride quality, rut resistance), cost, and experience when selecting a surface treatment. FIGURE 10 Influence of climate conditions on the selection of in-place recycling process. Percentages are based on the number of agencies and contractors with experience using the specific recycling process. TABLE 13 RANKING OF CLIMATES THAT CAN INFLUENCE THE CHOICE OF IN-PLACE RECYCLING PROCESSES Climate HIR CIR FDR Cold/Wet F G VG Hot/Wet G G VG Cold/Dry G VG VG Hot/Dry VG VG VG P = Poor, lower than 10% agencies and contractor with experience. F = Fair, between 10% and 25% agencies and contractors with experience. G = Good, between 25 and 50% agencies and contractors with experience. VG = Very good, greater than 50% agencies and contractors with experience. In summary, climate conditions need to be considered when selecting an in-place recycling process. Specific rea- sons for contractors’ and agencies’ climate preferences need to be defined in future research efforts. TABLE 14 STATE RESPONSES FOR TYPE OF SURFACE TREATMENT USED Surface Treatment: Indicate the typical top layer used for recycling projects. Surface Treatments Agency Responses HIR CIR FDR Fog and Chip Seal ID, KS, MT, NE CA, ID, IL, IA, MO, MT, NV, WA CA, GA, IL, IA, MT, SC, TX Fog Seal — — CA Structural Overlay AR, AZ, CO, FL, ID, IL, IA, KS, MD, MO, NC, NE, NY AZ, CA, CO, CT, DE, ID, IL, IA, KS, MN, MO, NC, ND, NE, NH, NV, NY, OR, RI, SD, UT, VA, VT, WA, WI, WY AZ, CA, CO, CT, DE, ID, IL, IA, MN, MO, NC, ND, NE, NH, NV, NY, OR, SC, SD, TX, UT, VA, WI, WY Integral Overlay AZ, CO, IA, KY, MO, TX CA, SD, VA, WA SD, VA Microsurfacing — CA, IL, UT CA, DE, IL, UT OGFC — NV NV Non-Structural Overlay AZ, CA, IL, KS, KY, NY, TX AZ, CA, IL, NY, VA, VT, AK AK, CA, GA, IL Slurry Seal ND, NE IL, WI IL, MD

13 FIGURE 11 Types of surface courses used with in-place recycling processes. Percentages are based on the number of agencies and contractors with experience using the specific recycling process. The agency and contractor responses were ranked and summarized to indicate currently used surface treatments for in-place recycling projects (Table 15). TABLE 15 SURFACE TREATMENT SELECTION Type of Surface Treatment HR CIR FDR Overlays Structural Often Frequently Frequently Non- Structural Sometimes Sometimes Sometimes Integral Sometimes Sometimes+ Sometimes*+ OGFC Rarely* Rarely Rarely Fog/Chip Sometimes* Sometimes Often* Microsurfacing Sometimes* Sometimes* Sometimes Slurry Sometimes* Rarely Rarely Fog Seal Rarely* Rarely Rarely Rarely = lower than 10% average of agency and contractor with experience. Sometimes = between 10% and 25% average of agency and contractor with experience. Often = between 25% and 50% average of agency and contractor with experience. Frequently = greater than 50% average of agency and contractor with experience. *Contractor response was significantly higher than agency with experience. +By definition, “integral” refers to HIR processes, however some state agencies indicated use on CIR ad FDR. The preference for using a structural overlay when struc- tural capacity improvement is not needed requires further research to define the criteria required to select this option. The ability of other surface treatments to provide acceptable surface courses needs to be explored. MATERIAL SELECTION AND MIX DESIGN The second step in the development of an in-place recycling project includes assessing the in-situ layer and reclaimed asphalt pavement (RAP) properties, selecting materials, and providing mix designs for setting the job mix formulas. The information varies on the basis of the recycling process (Figure 12). FIGURE 12 Work needed to select materials and establish job mix formulas. In-Situ Layer Properties In-situ properties are needed to evaluate the need for different designs for different segments of the project. The ability of the underlying layers to support the construction equipment and the variability in layer thicknesses that can affect a reason- able selection of milling depths also need to be determined. A number of approaches can be used to define the thickness and stiffness of the underlying layers. The most common methods of assessment include coring, boring logs for base and soils classifications, dynamic cone penetrometer testing (DCP), California bearing ratio, or resistance value (R-value) from soils testing or historical records, FWD layer modulus, ground-penetrating radar (GPR), or local experience (Jahren et al. 1999; Loizos and Papavasiliou 2006; Loizos 2007; Malick et al. 2007). The initial use for this testing is to • Determine the ability of the subgrade to support the weight of the recycling equipment, • Evaluate needs for increased structural capacity,

14 In summary, the availability or collection of in-place material properties information needs to be considered when developing the project design, specifications, and agency estimates of project costs. FIGURE 13 Methods of assessing in-place layer properties by agencies and contractors. Percentages are based on the number of agencies and contractors with experience using the specific recycling process. TABLE 17 SOURCES OF EXISTING IN-PLACE MATERIAL PROPERTIES (average of agency and contractor percentages) Layer Property Testing HIR CIR FDR Coring to Determine Thickness Often** Frequently* Frequently* Boring to determine thickness Sometimes Often Frequently FWD Sometimes Sometimes Often GPR Rarely Rarely Rarely Rarely = lower than 10% average of agency and contractor with experience. Sometimes = between 10% and 25% average of agency and contractor with experience. Often = between 25% and 50% average of agency and contractor with experience. Frequently = greater than 50% average of agency and contractor with experience. *Contractor response was significantly higher than agency with experience. **Agency response was significantly higher than contractor with experience. FIGURE 14 Example of comprehensive preconstruction testing program (based on original figure by Wirtgen 2004). • Provide information for the structural design (FDR), and • Identify sections in need of different treatments. States typically use the same preconstruction field test- ing, regardless of the recycling process (Table 16). Agency preferences for field-testing methods are borings and coring investigations. Fewer than a third of the agencies use FWD testing for determining the layer modulus and project vari- ability. Fewer than 9% of the states use GPR for preconstruc- tion project assessments. The use of GPR testing will likely increase as the technology becomes more widely available in the coming years because it can provide information quickly on layer thickness, presence of moisture, and sections of the project in need of different designs. TABLE 16 AGENCY TESTING FOR LAYER PROPERTIES Preconstruction Field Testing: Before construction, I typically use: Preconstruc- tion Work States HIR CIR FDR Coring to Determine Thickness AK, AL, CA, CO, DE, GA, ID, IA, MD, MN, NC, ND, NE, NV, OR, SC, SD, TX, UT, VA, VT, WY AR, AZ, CA, CO, FL, GA, ID, IA, KY, MD, MO, MT, NE, NH, NV, NY, OR, RI, SD, UT, VA, VT, WA, WY AK, AL, CA, CO, CT, DE, GA, ID, IA, MD, MN, MO, MT, NE, NH, NV, OR, SC, SD, TX, UT, VA, WY Boring to Check Depth of Base and HMA CO, ID, KS, MD, MO, MT, TX, WA CO, CT, DE, ID, KS, MD, MO, MT, NH, NV, OR, SD, UT, VA, WA, WI AR, AZ, CA, CO, FL, ID, IA, KS, KY, MD, MO, MT, NC, NE, NY, TX, VT, WA FWD Testing AR, AZ, CO, FL, ID, MD, NC, NE, TX, VT, WA AZ, ID, MD, MN, NC, NE, NV, OR, RI, SD, UT, VA, VT, WA, WY AK, AL, CA, ID, MD, MN, MT, NC, NE, NV, OR, SD, TX, UT, VA, VT GPR Testing MT, TX MN, MT MN, MT, TX, AK Fewer field tests are conducted on HIR projects by con- tractors and agencies compared with CIR and FDR projects (Figure 13). Core thickness, boring logs, and samples are the most commonly used preconstruction field tests. Contrac- tors are significantly more likely than agencies to conduct field tests for FDR projects. The average of the percentage of agencies and contrac- tors using a given method of assessing the in-place material properties was used to rank method preferences (Table 17). Wirtgen (2004) provides a suggested comprehensive testing plan for conducting a detailed investigation. The testing program includes cutting a test pit, coring, and DCP testing (Figure 14). A combination of the distress surveys and field tests provides the engineer with sufficient informa- tion to evaluate any design adjustments needed for various sections of roadway.

15 FIGURE 15 Comparison of testing for RAP properties between agencies and contractors. Percentages are based on the number of agencies and contractors with experience using the specific recycling process. TABLE 19 LABORATORY TESTING PROGRAMS Preconstruction Laboratory Testing HIR CIR FDR Binder Content Often Frequently* Often* Recovered Binder Properties Often Often* Sometimes* Gradations Often Frequently* Often* % Fines Often Frequently* Often* Application Rates Often Frequently* Often* Material Properties for Additives Often Frequently* Often* Rarely = lower than 10% average of agency and contractor with experience. Sometimes = between 10% and 25% average of agency and contractor with experience. Often = between 25% and 50% average of agency and contractor with experience. Frequently = greater than 50% average of agency and contractor with experience. *Contractor response was significantly higher than agency with experience. Preconstruction testing is key to designing recycling mixes. The time needed for this testing as well as the costs to the project need to be considered in developing cost esti- mates and project timelines. New Materials and Additives A range of new materials and additives can be used to pro- duce desired mix properties and early performance. New aggregates may be added to adjust the final gradation. New binders (paving-grade asphalts, emulsions) are used to soften aged asphalt in the RAP and provide more flex- ibility of the final asphalt concrete layer. Recycling agents and rejuvenators can be used instead of, or in conjunction with, the new binders to improve the binder performance properties. Although each material can be added individu- RAP Properties RAP binder content, RAP binder properties, and RAP aggregate gradation are needed for the appropriate selection of grades of new aggregates, new binders, recycling agents, and additives. The most common agency preconstruction laboratory testing focuses on RAP gradations and binder contents, material properties, and recovered binder proper- ties (Table 18). The same testing program is used regardless of the recycling process; however, agencies are more likely to test the recovered binder properties for HIR projects than for either CIR or FDR projects. TABLE 18 AGENCY RESPONSES FOR PRECONSTRUCTION LABORATORY TESTING FOR IN-PLACE RECYCLING PROJECTS Preconstruction Laboratory Testing: Before construction, I typically determine: Laboratory Testing States HIR CIR FDR Aggregate Gradations of Cores or Millings AR, CA, FL, ID, KS, KY, MF, NC, NE, TX, WA CA, CT, DE, ID, KS, MD, MN, ND, NH, NY, RI, SD, UT, VT, WY AK, AL, CA, DE, GA, IA, MD, MN, SD, UT, VT Application Rates of Bind- ers or Other Additives CA, CO, IA, KS, MD, NE, NY, TX, WA CA, CO, CT, IA, KS, MD, MN, NE, NH, NY, SD, UT, VT, WY AL, CA, CO, GA, ID, IA, MD, MN, NE, SC, UT, VT, WY Binder Content of Cores or Millings AR, CA, FL, ID, KS, KY, MD, NC, NE, TX, WA CA, CO, CT, ID, KS, MD, MN, ND, NE, NH, NY, WY AL, CA, GA, MD, MN, WA Material Properties of Any Liquids, Stabilizers, Rejuvenators, Additives, or Admixtures to Be Added CA, CO, FL, GA, IA, KS, MD, NC, NE, NY, TX, WA CA, CO, IA, KS, MD, NE, SD, UT, WY AL, CA, CO, GA, IA, MD, NE, TX, UT, VT, WY Percent Fines of Millings CA, FL, ID, KY, MD, NC, NE, TX CA, ID, MD, UT, VT, WY AK, CA, ID, IA, MD, NE, UT Recovered Binder Proper- ties from Cores or Millings CA, FL, GA, KY, MD, NC, NY, TX, VT, WA CA, MD, WY AK, CA, MD Contractors conduct more tests before construction than the agencies (Figure 15). Contractors and agencies tend to agree more often on testing of HIR than on either CIR or FDR projects. The agency and contractor responses were used to rank and summarize current practices for laboratory testing (Table 19).

16 • Emulsions – CSS-1, CSS-1h, and CSS-1hP – CMS-2S – HFMS-2, HFMS-2S – HF-150, HF-300P – Proprietary solventless emulsions • Asphalt binders in fresh mix – Performance-graded asphalt, softer grades – Viscosity-graded asphalts (e.g., AC 10, AC 20) – Foamed asphalt Emulsions are a combination of small asphalt globules suspended in water by the use of surfactants. A sample of the emulsion grades used in recycling projects is shown in Table 20. Regardless of the source of the emulsion specification, three groups of material property tests are typically needed to determine the properties of emul- sions (water, asphalt, additives), distillation of emulsions (removal of water), and the recovered base asphalt (resi- ally, it is common practice to introduce new aggregates and asphalt by adding new HMA to the recycled materials. Additives and stabilizers can be added to improve stiff- ness, moisture resistance, and rut resistance; reduce ravel- ing; help dry moist RAP and soils; and control the rate of set of emulsions. New Aggregates When new aggregates are added, existing aggregate grades are typically used such as 3/8 in. minus, no. 57 stone, ½ in. minus sizes. Standard aggregate gradations locally available are typically used when gradations of the final mix need to be adjusted. Asphalt Binders Asphalt binders used in recycling processes can be typical paving-grade asphalts or emulsions: TABLE 20 REQUIREMENTS FOR CATIONIC EMULSIFIED ASPHALTS (based on ASTM D2387-05; D977-05; Oregon DOT 2010) Type Medium Setting Slow Setting Grade HFMS-2 HFMS-2s HF-150 CMS-2S CSS-1 CSS-1h Min Max Min. Max Min Max Min Max Min Max Min. Max Tests on Emulsions Viscosity, Saybolt Furol at 25ºC (77ºF) SFS 100 50 35 150 20 100 20 100 Viscosity, Saybolt Furol at 50ºC (122ºF) SFS 100 450 Storage Stability Test, 24-h, % 1 1 1.5 1 1 1 Demulsibility, 35 mlL, 0.8% Dioctyl Sodium Sulfonsuccinate, % 40 Coating Ability and Water Resistance Coating , Dry Aggregate good good good Coating, After Spraying fair fair fair Coating, Wet Aggregate fair fair fair Coating, After Spraying fair fair fair Particle Charge Test positive positive positive Sieve Test, % 0.1 0.1 0.1 0.1 0.1 0.1 Cement Mixing Test, % 2 2 Tests on Distillation Oil Distillate, by Volume of Emulsions, % 0.5 4 12 Residue, % 65 65 62 60 57 57 Tests on residue from distillation test Penetration, 25ºC (77ºF), 100g, 5 s 100 200 200 150 250 100 250 100 250 40 90 Ductility, 25ºC (77ºF), 5 cm/min, cm 40 40 40 40 40 Solubility in Trichloroethylene, % 97.5 97.5 97.5 97.5 97.5 97.5 Float Test, 60ºC (140ºF), s 1,200 1,200 1,200

17 due). Traditional emulsion specifications use one or more tests to define the asphalt residue properties: absolute and kinematic viscosity, penetration, and ductility testing. The existing specifications rely on older methods of grading asphalts (e.g., penetration grades, viscosity grades); how- ever, most states now specify asphalt products using the performance grading (PG) specifications. Clyne et al. (2003) explored PG specification testing (AASHTO MP1) to classify the asphalt residue from three emulsions used in the same region of Minnesota (Table 21). Based on these results, the engineered emulsion (EE) and HFMS-2P would be expected to remain more flexible at colder temperatures than CSS-1. Both CSS-1 and HFMS-2P would be expected to be less sensitive to movement under traffic at summer temperatures than the EE product. This is also supported by research for Federal Lands (Johnston and King 2008). Research by Epps et al. (2001) suggested that emulsion specifications use the concept of PG binder prop- erties so that recycling binders can be selected for project- specific environmental conditions. TABLE 21 PG GRADING FOR MINNESOTA EMULSIONS (based on Clyne et al. 2003) Emulsion Performance Grade CSS-1 PG 52-28 Engineered Emulsion (EE) PG 46-34 HFMS-2P PG 52-34 Emulsions historically used in the same environmental conditions may have base asphalts with a wide range of performance-graded asphalt properties that will likely influ- ence the success or failure of recycling projects. Emulsion specifications need to be updated so that users can select binders on the basis of performance properties. Asphalts—Paving Grades Paving-grade asphalt can be specified by the standard PG specification by using the desired properties of the com- bined asphalt (i.e., combination of new and RAP asphalt). A formal blending program can be conducted to select the fresh binder PG specification, or a less formal “bumping” down one grade to account for the stiffening of the fresh binder because the aged RAP binder can be used. The Texas Department of Transportation is conducting research into the use of Superpave® PG specifications for asphalts used in near-surface applications. Specific guidance is needed for using PG specifications for recycling project asphalts. Asphalt—Foamed Asphalt One method of adding fresh binder to cold recycling pro- cesses (CIR and FDR) is to foam it during the mixing pro- cess in the recycling train. Foamed asphalt is produced by injecting a small amount of water into hot asphalt as it is mixed with the recycled materials (Figure 16). As the hot liquid and water mix, the liquid expands as the water turns to steam, creating a thin film of asphalt with about 10 times more coating potential. Foaming facilitates better dispersion of the asphalt into the materials to be recycled. The two key parameters that control the quality of the foam are the: FIGURE 16 Foamed asphalt process during construction (based on original figure by Wirtgen, 2004). • Expansion ratio (minimum of 10 times; Wirtgen 2004) and • Half-life of the foam (minimum of 8 s; Wirtgen 2004). The expansion ratio is defined as the ratio between the maximum achieved volumes of the foam to its original vol- ume. The half-life is defined as the time elapsed from the time the foam was at the maximum volume to the time it reaches half of the maximum volume. Larger expansions and longer half-life are considered desirable properties for foamed asphalt. Marquis et al. (2002) noted that the quality of foamed asphalt mix is strongly related to the quality of the foam as measured by the expansion ratio and the half-life. Optimum settings for the foaming process need to be set in the mix design phase on the basis of parameters needed to produce peaks in stiffness or modulus of the mix. It should be noted that there need to be between 8% and 20% fines in the FDR to achieve the desired results for foamed asphalts, although 100% RAP mixes can be prepared with lower percentages of fines (Matthews 2008). The Wirtgen (2004) manual recom- mends a range of gradations suitable for foamed asphalt (Fig- ure 17). One of the main advantages to using foamed asphalt

18 cal hourly HMA plant production, and plant operators will be reluctant to change plant operations for small orders of HMA. When they are willing to change mixes, the cost of the specialty mix will be substantially higher than typical HMA, and the availability will depend on the plant’s abil- ity to interrupt its production schedule. The most economi- cal and practical approach is to start with the properties of the typical new HMA that is locally available and adjust the RAP gradation if at all possible. Recycling Agents and Rejuvenators Recycling agents (RAs) are used to restore the aged asphalt to the desired binder properties. ASTM D4552 (Table 22) classi- fies petroleum product additives specifically for hot mix recy- cling methods. The RA classifications are viscosity graded, with the lower the number designation representing the low- est viscosity. Products meeting the RA 1 through RA 75 des- ignations are typically used for recycled mixes with more than 70% RAP in the mixes. When more than 30% of new aggregate is used, the RA 250 and RA 500 grades are more appropriate. The Pacific Coast Conference on Asphalt Speci- fications defines RAs as hydrocarbon products with physical characteristics selected to restore the aged asphalt binder to the current asphalt binder specifications. By this definition, a softer grade of asphalt can be classified as an RA. ASTM D5505 provides specifications for emulsified recycling (ER) agents (Table 23). The base asphalt in these products increases in stiffness (viscosity) with increases in the grade number. The ER-1 is a petroleum derivative com- patible with asphalts. Its main function is to rejuvenate aged asphalt. The ER-1 material is viscosity graded, and there are no requirements for viscosity measurements on the residue after rolling thin film oven (RTFO) testing. The ER-2 and ER-3 grades are a combination of rejuvenators and asphalt components. These ER agents are typically used when the recycled HMA needs additional asphalt (e.g., when adding new aggregate). They are considered a penetration-graded instead of emulsions is that the roadway can be opened to traffic immediately after compaction because no curing time is needed (Lane and Kazmierowski 2005a). If the fines con- tent of the RAP is insufficient, then additional mineral filler will be needed (see the section on additives and stabilizers). FIGURE 17 Suggested gradation range for foamed asphalt (based on original figure by Wirtgen 2004). New HMA Fresh mix is commonly used in HIR, but it is used very rarely in CIR and not at all in FDR. The new HMA is used to easily adjust the gradation of the HIR. The gradation of the new HMA is selected to achieve the desired final grada- tion, and the binder grade of the new HMA is selected to help soften and rejuvenate the RAP binder. Mixes reported as used include minus 1.9 mm (¾ in.) dense-graded HMA, fine-graded HMA, minus 12.5 mm (1/2 in.) open-graded HMA, and stone matrix asphalts. The simplest means of obtaining economical and timely new HMA supplies is to use HMA mixes typically produced by the plant supplying the mix. The amount of new HMA needed for a recycling job is small compared with the typi- TABLE 22 ASTM D4552 CLASSIFICATIONS FOR HOT MIX RECYCLING AGENTS (ASTM D4552 2009) Test ASTM Test Method RA 1 RA 5 RA 25 RA 75 RA 250 RA 500 Min Max Min Max Min Max Min Max Min Max Min Max Viscosity at 140°F, cSt D 2170 or D 2171 50 175 176 900 901 4,500 4,501 12,500 12,501 37,500 37,501 60,000 Flash Point, COC, °F D92 425 — 425 — 425 — 425 — 425 — 425 — Saturates, wt% D 2007 — 30 — 30 — 30 — 30 — 30 — 30 Tests on residue from RTFO or TFO oven 325°F (D 2872 or D 1754) Viscosity Ratio — — 3 — 3 — 3 — 3 — 3 — 3 Wt Change ± % — — 4 — 4 — 3 — 3 — 3 — 3 Specific Gravity D 70 or D 1298 Report Report Report Report Report Report

19 material because the penetration is used to set limits on the residue after RTFO conditioning. In addition to the ASTM recycling agents and ASTM ER agents, some states may include a state-developed specifi- cation such as the one from Kansas (Table 24). There are also proprietary recycling products on the market, such as engineered emulsions specifically designed to address dis- advantages of conventional recycling agents, in particular in-place recycling methods. Proprietary products specifi- cally designed for in-place recycling that have been used by agencies and contractors are • CIR-EE, • Reflex, • Fortress, • Pass-R, • ERA-25, • ARA-1P, and • Reclamite. Additives and Stabilizers Geiger et al. (2007) summarized the reasons for using sta- bilization to improve the characteristics of base materials as • Reducing plasticity index, • Reducing swelling potential of the in-situ soils, • Increasing base durability and strength, • Reducing dust during construction, • Waterproofing the in-situ soils, • Drying wet in-situ soils, • Conserving natural resources (aggregates), • Reducing construction costs, and • Providing a temporary wearing surface. TABLE 24 ASPHALT REJUVENATING AGENT (based on Kansas specification 1205) Property Requirement Viscosity, Saybolt–Furol at 25ºC, s 15–100 Residue, % min. 60 Sieve Test, % max. 0.1 Oil Distillate, % max. 2 Storage Stability, 24 h, % max. 1 Tests on Residue from Distillation Asphaltenes, % max. 15 Penetration @ 4°C, 100g, 5 sec. 150–250 The types of additive(s) used with CIR processes are based on the desired mix property improvements, such as improved stripping resistance, rut resistance, layer stiffness for higher traffic levels, controlled rate of set of emulsions, minimized raveling until the wear course is placed, and additional fines TABLE 23 ASTM D5505 SPECIFICATIONS FOR EMULSIFYING RECYCLING AGENTS (ASTM 2009) Tests Test Method ER-1 ER-2 ER-3 Min Max Min Max Min Max Testing on emulsion Viscosity, 50°C, SSF D224 100 20 450 20 450 Sieve, % D6933 0.1 0.1 0.1 Storage Stability, 24 h, % D6930 1.5 1.5 1.5 Residue, by Distillation, % D6997 65 65 65 Dilution — ReportA Specific Gravity D70 Report Report Report CompactibilityB varies Report Report Report Testing on residue from distillation Viscosity, 60°C, cSt D2170 50 200 30 30 Saturates, % D2007 30 Solubility in Trichloroethylene D2042 97.5 97.5 97.5 On residue from distillation after RTFOC Penetration, 4°C, 50 g 5 s D5 75 200 5 75 RTFO, Weight Change, % D2872 4 4 4 Notes: AER-1 shall be certified for dilution with potable water. BThis specification allows a variety of emulsions, including high-float and cationic emulsions. The engineer should take the steps necessary to keep incompatible materials from co-mingling in tanks or other vessels. It would be prudent to have the chemical nature (flat test for high-float emulsions, particle charge test for cationic emulsions, or other tests as necessary) certified by the supplier. CRTFO shall be the standard. When approved by the engineer the Thin Film Oven Test (Test Method D 1754) may be substituted for compliance testing.

20 needed to meet the desired gradation. FDR materials can be stabilized with most of the additives used for HIR and CIR improvements. Stabilization improves the load-bearing quali- ties of the mostly unbound pulverized materials and is classi- fied by how it improves base properties (ARRA 2001): • Mechanical, • Chemical, • Bituminous, and • Combinations. Mechanical stabilization is developed by using par- ticle interlock typically achieved by pulverizing RAP and base materials and then compacting to the desired density. Because all of the recycling methods include compaction, mechanical stabilization can be considered a secondary sta- bilizing mechanism for all of the methods. Chemical stabilization mixes the pulverized RAP and base or subgrade materials with cementitious materials such as calcium chloride, magnesium chloride, lime (hydrated or quicklime), fly ash (Class F or C), kiln dust (cement or lime), portland cement, or other chemicals (ARRA 2001). Some of these chemical stabilizers can be added either dry or in slurry form. Bituminous stabilization uses an asphalt emulsion, ER agent, or foamed (expanded) asphalt. It is not unusual to see combinations of stabilizers such as fly ash and asphalt emulsion or fly ash and portland cement. Combinations of stabilizing methods and additives are commonly used to improve properties. Liquid calcium chloride is used to improve freeze/thaw resistance by lowering the freezing point of reclaimed base material. The stiffness of the base is improved by the bond- ing of the soil and RAP particles. The first application of the liquid is blended with the pulverized material; the stabilized base is shaped and graded and then sealed with a second application of calcium chloride. Portland cement is used to increase compressive strengths of bases by providing a cementitious bonding of the soil and RAP particles. Portland cement works best with a plasticity index of less than about 10 (Matthews 2008; Thompson et al. 2009) and fewer than 10% fines (Franco et al. 2009). Higher percentages of fines can be tolerated while still improving the load-carrying capability of the soil. Cement-stabilized bases continue to slowly gain strength over time and work best with granular materials with low plasticity. Another advantage to using cement as a stabilizer is that excess mois- ture can be quickly removed from the pulverized material. One disadvantage is when used as a stabilizer, the recycled material has a tendency to show shrinkage cracking. Lime (calcium hydroxide) works best when there is reac- tive clay in base materials, as lime reduces the plasticity of the clay materials. Lime is typically used when the plasticity index (PI) is greater than 8 (Matthews 2008) and fines con- tents are greater than 10% (Franco et al. 2009). Thompson et al. (2009) recommend using 1% hydrated lime when the PI is between 10 and 16 and 2% when the PI is greater than 16. The reduction in plasticity helps minimize swelling, reduce moisture damage, and improve the base strength. Like port- land cement, lime can help reduce initial excess moisture in the pulverized base materials. Too much lime can result in shrinkage cracking. Quicklime (calcium oxide) reacts with water to form cal- cium hydroxide, a reaction that generates heat, and the solid nearly doubles in volume. Because of the fast reaction of the quicklime, it is used for set control or early strength gains. The benefits are the same as using hydrated lime. Fly ash, a pozzolanic material, also provides improved base strength through a cementitious bonding of the particles when in the presence of water. Moisture resistance is improved by a reduction in the permeability of the base materials. Asphalt emulsions, a mixture of asphalt cement, water, and an emulsifying agent, improve the strength and mois- ture resistance of the base material, soften the aged asphalt binder in the RAP, and reduce shrinkage cracking seen with cement and lime stabilizers. When the emulsion breaks, the asphalt droplets join, and the water separates from the asphalt. Compaction helps force the water out of the stabi- lized base, but sufficient time for the moisture content to drop below about 1.0% is still needed for all of the moisture to evaporate before the placement of the next lift. Combinations of additives and stabilizers have been used with asphalt binders to improve properties of the final prod- uct. For example, Naizi and Jalili (2009) found that using emulsions with lime slurry or portland cement improved moisture resistance and increased both the final mix stiffness and indirect tensile strength. Thomas et al. (2000) evaluated the combination of fly ash and lime, which showed improved mix stiffness but promoted shrinkage cracking. A combina- tion of EE and lime slurry provided improved flexibility at cold temperatures and minimized shrinkage cracking. The Wirtgen (2004) manual notes that cement is routinely used with bitumen emulsions to improve moisture resistance, tensile strength, fatigue resistance, and retained strengths. Cement and emulsion combinations need less curing time before traffic can be permitted on the recycled surface. Information on the use of typical additives and stabilizers compiled from the survey responses and from the literature is summarized in Table 25.

21 FIGURE 18 Basic steps in recycled mix designs (based on FHWA 1997). TABLE 25 SUMMARY OF USES FOR ADDITIVES AND STABILIZERS IN RECYCLING PROCESSES Additive or Stabilizer CIR FDR Moisture Resistance Freeze/ Thaw Resistance Rut Resis- tance Layer Stiffness Rate of Set Control Minimize Raveling Mechanical Stabilization Chemical or Bituminous Stabilization Calcium Chloride X X X Portland Cement X X X X X X Lime X X X X X Quicklime X X X X X Fly Ash X X X Limestone Fines X X X X Fibers X X X Asphalt X X X X X X Recycling Agents X X X X X In summary, additives and stabilizers need to be selected on the basis of their ability to improve key material and mix properties. Mix Designs Regardless of which recycling process is used on a project, the steps in the mix design process are similar (Figure 18). New and RAP Binder, RA Selection Once the gradation blend of RAP and new aggregate is determined, the binder grade, quantity, and any recycling or rejuvenating agent need to be identified. For HIR, this can be done by the use of blending charts, which can be adapted for viscosity or Superpave PG binder tests (Figure 19). The viscosity or G*/sinδ for the RAP binder is plotted on the left y-axis and the properties of the new asphalt, or RA for CIR, are plotted on the right. A line is drawn horizontally across the graph, from left to right, until it intersects the diagonal viscosity line. The percentage of new asphalt or RA needed is read off the bottom horizontal axis. More comprehensive selection methods will blend the anticipated percentages of RAP, and new binder will use the full Superpave binder property to select the new binder grade. Mix Design Methods The most commonly used mix design methods vary by the in-place recycling process. HIR mix designs are usually based on standard HMA mix design methods. CIR and FDR are based on emulsion or foamed asphalt methods, which include

22 FIGURE 19 Blending chart used to select the percent of new asphalt or additive needed to provide the desired binder properties (based on FHWA 1997). • EEs – Caltrans: 75 blow Marshall – Iowa DOT: 4-in. gyratory with 30 revolutions – SemMaterials: 6-in. gyratory with 30 gyrations • Emulsions – Wirtgen: 75 blow Marshall – Ontario Ministry of Transportation (MTO): 75 blow Marshall • Foamed asphalt – Iowa DOT: 4-in. gyratory with 25 revolutions – Wirtgen: 75 blow Marshall – Ontario MTO: 75 blow Marshall Because mix designs are intended to represent field condi- tions, curing periods before testing are included in emulsion (engineered or traditional) and foamed asphalt mix designs. As with the compaction methods, each mix design method varies in its curing procedures (Thompson et al. 2009): • EEs – Caltrans: Cure at 140ºF to constant weight – Iowa DOT: 48 h at 140ºF – SemMaterials: 72 h at 140ºF • Emulsions – Wirtgen: 72 h at 104ºF. For high traffic (i.e., greater than 5 million ESALs), the specimens are com- pacted at the anticipated final field moisture content and cured in sealed containers for 40 h at 104ºF. – Ontario MTO : 48 h at 140ºF, soaked for 24 h at 77ºF, or vacuum saturated for 60 min at mmHg pressure. • Foamed asphalt – Iowa DOT: 72 h at 105ºF – Wirtgen: Same as for emulsions – Ontario MTO: Same as for emulsions Once the specimens have cured, various properties of the specimens are determined: • EEs – Caltrans: Marshall stability at 104ºF • 1,250 lb minimum dry stability • 70% minimum retained strength ratio – Iowa DOT: Marshall stability at 100ºF • 1,000 lb minimum stability – SemMaterials: Indirect tensile strength, resilient modulus, and modified cohesiometer • 35 to 40 psi minimum (dry) • 20 to 25 psi minimum (wet) • 70% minimum ratio • 120 to 150 ksi minimum for resilient modulus • Emulsions and foamed – Ontario MTO: Indirect tensile strength (dry and wet), retained strength ratio • 50 psi minimum (dry) • 25 psi minimum (wet) • 50% minimum for ratio The most common approach by state agencies in designing recycled mixes is to do nothing (Table 26). When agencies do mix designs, either the Superpave or Marshall methods are commonly used. None of the agency respondents indicated that they use the standard Hveem method, and only four states use the Wirtgen (2004) method for CIR or FDR. For states indicating “other,” the design methods mentioned were the Portland Cement Association (PCA) soil–cement mix design (PCA 2005), Proctor method (optimum dry density and mois- ture content), modified Proctor (Kim and Labuz 2007), and unconfined compressive strengths (geotechnical testing). TABLE 26 AGENCY AND CONTRACTOR RESPONSES TO MIX DESIGN METHODS Mix Design Testing: Before construction, I or my contractor design our recycled mixes based on the following method: Mix Design Methods Agency Responses HIR CIR FDR Do Not Do Mix Designs CA, ID, IA, MO, VT, WA CA, DE, ID, IA, NC, NH, NV, RI, SD, VT, WA, WI CT, DE, ID, MN, MT, NC, NH, NV, NY, SD, VT, WI Hveem — — — Marshall AZ, KY, NE AZ, MN, NE, OR, VA, WY VA Superpave CO, KS, MO, ND, UT, VT CO, KS, MO, ND, UT, VA MD, MO, UT, VA Wirtgen — VA AK, CA, IA, VA Other NY, TX CT, MT, NY AL, CO, GA, NE, NY, SD, WY Between 20% and 42% of the states do not develop mix designs for recycling projects (Figure 20). Comments

23 by the states indicated that they require the contractor to supply the designs. Contractors typically design using the Marshall or Superpave mix designs. Contractors are more likely to use either Marshall or Wirtgen or no mix design at all for FDR designs. FIGURE 20 Comparison of mix design methods used by agencies and contractors. Percentages are based on the number of agencies and contractors with experience using the specific recycling process. AASHTO, the Associated General Contractors of Amer- ica, and the American Road and Transportation Builders Association Joint Committee Task Force 38 adapted Mar- shall (50 blow) and Hveem mix designs (ARRA 2001) for use with recycled mixes. Since the Task Force report came out, several researchers have evaluated the suggested designs, primarily the 50 blow Marshall method, which con- sists of two parts: determination of optimum water content and determination of optimum binder content. Two stud- ies (Salomon and Newcomb 2000; Lee et al. 2002) evalu- ated this CIR mix design method and noted the following disadvantages: • Time needed to complete mix design is 8 days. • Information on when new aggregate should be added to the mix (i.e., no suggested gradation bands) was missing. • Time needed for emulsion to break was not considered in sample preparation. • Heating time for emulsion was not specified. • Temperature differences for different emulsions were not addressed. • Applicability of using standard HMA testing for bulk specific gravity (i.e., direct immersion of high air void mixes in water) was not addressed, but should be con- sidered because of the high air voids in recycling mixes. • Specific procedures for determining optimum values of water and emulsions were not clearly defined. Preparation, mixing (order of addition), mixing tempera- tures, curing times (before and after compaction), curing tem- peratures, and determination of specific gravities are the focus of a number of current agency and academic research projects. Recent published results indicate that gyratory compac- tion is useful for preparing samples for all of the recycling processes. In particular, using gyratory compaction for FDR seems to provide a compacted density closer to actual in-situ densities than the other methods. Mallick et al. (2002) and Kim and Labuz (2007) found 50 gyrations produced labora- tory-compacted samples with densities similar to those found in the field projects. Other researchers investigated using 30 gyrations for preparing FDR samples (Cross 2002; Lee and Kim 2007b; Thompson et al. 2009). Some concern was expressed about the need to provide drainage for the water pressed out of the CIR and FDR mixes (Mallick et al. 2007), and a slotted gyratory mold was used when compacting these mixes. Gyratory compaction can also be used for foamed asphalt samples (Kim and Lee 2006; Kim et al. 2007b). The load-carrying capability of the recycled mix is evalu- ated with Marshall or Hveem stabilities. The indirect ten- sile strength test (IDT) is also used either in place of, or in addition to, the stabilities. Both dry and wet IDTs are used by a number of agencies and contractors to determine the moisture sensitivity by evaluating the retained strengths (i.e., tensile strength ratio, TSR). States that evaluate rutting potential with loaded wheel testers (i.e., asphalt pavement analyzer, Hamburg) for their HMA mixes also use these tests for the recycled mixes. The PCA (2005) and general unconfined compressive strength approaches to design recommend a range of com- pressive strengths at various times after curing. For example, the PCA method uses limits for strengths between 2.07 to 2.76 MPa (300 and 400 psi) at 7 days. Franco et al. (2009) recommend including the determination of the Atterberg limits in the mix design methods for FDR. The FDR mix design method used depends on the type of stabilizer. Because FDR is essentially a method of produc- ing a stabilized base material, typical geotechnical tests are commonly used by agencies. These include using a Proc- tor or modified Proctor determination of optimum moisture content and maximum dry density. Strength testing is con- ducted using unconfined compressive strength, California bearing ratio, or R-value tests. When cement is used, the PCA method for soil–cement stabilization may be used (PCA 2005). Other stabilized base mix designs for fly ash and lime stabilization can be used for CIR and FDR mixes. Combinations of additives and stabilizers such as emulsions and cement can be designed with CIR mix designs or by using geotechnical tests. Regardless of the mix design method used, there was gen- eral agreement that there is a lack of established curing times, temperatures, or humidity conditions. There is some agree-

24 ment that gyratory compaction for FDR samples is appropri- ate; however, the use of this compaction device for CIR mixes that still have significant water content needs to be evaluated. The average of the agency and contractor responses was used to rank and summarize the types of mix designs cur- rently used (Table 27). TABLE 27 MIX DESIGN PRACTICES Mix Design Methods HIR CIR FDR No Formal Design Sometimes Sometimes** Often** Superpave Rarely Rarely Rarely Marshall Sometimes Often Often* Hveem Sometimes Often* Sometimes Wirtgen Rarely Sometimes* Sometimes* Rarely = lower than 10% average of agency and contractor with experience. Sometimes = between 10% and 25% average of agency and contractor with experience. Often = between 25% and 50% average of agency and contractor with experience. Frequently = greater than 50% average of agency and contractor with experience. *Contractor response was significantly higher than agency with experience. **Agency response was significantly higher than contractor with experience. Superpave mix design methods need to be developed for designing recycled mixes. Curing times, temperatures, and humidity need to be standardized for CIR and FDR regard- less of the type of compaction used to prepare the samples. STRUCTURAL DESIGNS Structural design methods rely on the assessment of remain- ing pavement life and the needed structural changes for future traffic. Layer properties, thickness, and distress information are needed to determine the appropriate changes during main- tenance and rehabilitation activities. In some cases, neither the agency nor contractor assesses the structural capacity (i.e., the “No” answer option to the question in Step 3). In this case, the process moves directly to construction (chapter three). Established structural coefficients for the traditional AASHTO design are the most commonly used design approach by the agencies, followed by the use of FWD test- ing for layer properties (Table 28). Specific information on the value(s) of structural design coefficients was not col- lected in this survey; however, several values commonly used were found in the literature. A coefficient of 0.44 was recommended for HIR layers (In-Place Recycling Confer- ence 2008). For CIR materials, coefficients of between 0.25 and 0.28 were recommended by Kansas, 0.26 by Nevada, and 0.35 in a NCHRP Report 224 (Harrington 2008). Roma- noschi et al. (2004) recommended a coefficient of 0.18 for foamed asphalt-stabilized FDR. The Ontario MTO uses 0.20 to 0.28 (estimated from gravel equivalent) for foamed FDR (Thompson et al. 2009). TABLE 28 STATE AND CONTRACTOR RESPONSES TO STRUCTURAL DESIGN CONSIDERATIONS Structural Design: During project development, I consider the structural capacity of the recycled layer using: Structural Design Considerations Agency Responses HIR CIR FDR Established Structural Coefficients AZ, CO, IA, MT, NE, UT, WA AZ, CO, IA, MN, MT, NE, NV, OR, RI, SD, UT, WA, WI, WY AL, CO, IA, MN, NE, NV, OR, SD, UT, WI FWD AR, AZ, ID, MD, NE, TX, UT, VT, WA AZ, ID, ND, NE, NV, OR, SD, UT, VA, VT, WA AK, AL, CA, ID, MD, MT, ND, NE, NV, OR, SD, TX, UT, VA, VT Pre-Determined Layer Thickness AZ, FL, NC, WA AZ, CA, DE, NC, NV, SD, WA CA, DE, MD, NC, NV, SC, SD Laboratory Resilient Modulus MD VA MD, NE, VA Wirtgen (2004) uses a nomograph to estimate the layer coefficient, other layer properties, and anticipated amount of foamed asphalt (Figure 21). The nomograph is used by iden- tifying a given property and then moving vertically up or down to obtain estimates for the other layer properties. For example, given a structural number coefficient of 0.16 after stabilization, the anticipated initial stiffness would be about 750 MPa, a steady stiffness of 450 MPa, and indirect tensile strength of 150 kPa when using about 4% foamed asphalt for a range of AASHTO soil classifications. The advantage to the graph is that material properties are tied to the selection of the coefficients. This graph can also be used to estimate material properties for use with newer mechanistic–empiri- cal design methods. Alternatively, FWD testing to deter- mine the existing pavement layer stiffness (i.e., modulus) can be used to estimate structural coefficients. A number of agencies simplify their design process by using predetermined thicknesses for each of the recycling methods. Only a limited number of states use laboratory resilient moduli values for their designs.

25 written responses include the use of compressive strength, distress level, and R-value. The average of the agency and contractor responses was used to rank and summarize the current use of structural design approaches for in-place recycling (Table 29). TABLE 29 STRUCTURAL DESIGN APPROACHES Design Information Used HIR CIR FDR Structural Coefficient Sometimes Frequently Frequently FWD Sometimes Often Often Set Thickness Sometimes Often Often* Lab Resilient Modulus Rarely Sometimes Often* Rarely = lower than 10% average of agency and contractor with experience. Sometimes = between 10% and 25% average of agency and contractor with experience. Often = between 25% and 50% average of agency and contractor with experience. Frequently = greater than 50% average of agency and contractor with experience. * Contractor response was significantly higher than agency with experience. Information regarding structural coefficients and layer stiffness is needed for structural design considerations. These design parameters need to be agreed on before con- struction so that the final product meets or exceeds the desired performance. FIGURE 21 Relationships between layer properties and layer coefficients (based on original figure by Wirtgen 2004). Contractors are most likely to use structural coefficients and FWD values for structural design considerations (Fig- ure 22). Contractors are more likely than agencies to use either set recycled mix thickness or mix stiffness (resilient modulus) information for their designs. Differences between contractors and agencies are more noticeable in their choice of FDR designs. Other design considerations noted in the FIGURE 22 Information used for structural design approaches by agencies and contractors. Percentages are based on the number of agencies and contractors with experience using the specific recycling process.