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.
76 â¢ 2005â2010âAdditional performance (rut testing) and durability (permeability) testing is added for mix design approval. â¢ 2012âPercentage of RAP asphalt considered useful in mixtures is reduced from 100% to 75% and the original optimum mix design asphalt content is increased by add- ing additional virgin asphalt that is calculated as the per- centage of nonuseful RAP asphalt. Performance testing is conducted on samples prepared at the increased virgin asphalt level. â¢ 2015âHigh RAP (>25% RAP) mixtures can be used in any pavement lift and mix designs with 30% RAP are routinely approved. High RAP surface coarse mixtures were placed in 2012 and at the beginning of 2015. GDOT started using the Superpave mix design methodol- ogy during 1998, using four gyratory compaction levels: 50, 75, 100, and 125. The initial Superpave mix designs tended to produce coarser mixtures with lower optimum asphalt contents to resist rutting. At the same time, the mix design methodology changed; approximately 10% RAP was used in GDOT mixture. Between Superpave implementation in 1998 and 2005, the percentage of RAP gradually increased from 10% to 25%. By 2005, feedback from the GDOT maintenance division noted concerns with increased evidence of early pavement distresses on projects that used around 25% RAP and were less than 3 years old. The documented problems included permeable areas of the pavements (i.e., low density, which allows water to infiltrate) leading to increased moisture dam- age, more frequent evidence of segregation followed by seg- regation-related moisture damage, visible coarse streaking in the freshly placed mat surface, and a generally dry, aged look within a short period of time (Figure 26). All of these in-place problems with early pavement dis- tresses can be linked to inadequate asphalt film thickness (low asphalt content), which is the likely reason for the âdryâ look and the mixture is more: â¢ Difficult to handle; â¢ Likely to segregate; â¢ Difficult for the mixture to move uniformly as it is trans- ferred from the silo into the haul truck, from the truck into the paver, through the paver, and across the back of the screed; This chapter presents examples from those agencies that provide additional information for five topics: 1. Georgia Department of Transportation (GDOT) specifi- cation development that encourages the routine contrac- tor submittals of high RAP mixtures. 2. Contractorsâ perspectives for routinely produced high RAP mixtures for GDOT, as well as mixtures for other clients in surrounding states that use RAS and/or a combination of RAP and RAS. 3. Contractor suggestions for producing and placing RAS asphalt mixtures. 4. Locating and using county databases increases pave- ment performance evaluation information. 5. Evaluating the amount of recycled material asphalt transfer to the virgin aggregate during dry mixing at the plant (i.e., the time before the liquid virgin asphalt is added). CASE EXAMPLE NO. 1: GEORGIA DEPARTMENT OF TRANSPORTATION SPECIFICATION DEVELOPMENT FOR HIGH RAP MIXTURES This section describes GDOTâs implementation and refine- ment of specifications for high RAP mixtures that resulted in contractors routinely submitting mix designs using from 30% to 40% RAP in any pavement layer. A timeline sum- mary of GDOTâs specification implementation and refine- ment follows: â¢ 1998âImplement Superpave mix design methodology. â¢ 1998â2005âPercentage of RAP used in Georgia mix- tures increases from 10% to 25% and a variety of early pavement distresses associated with low asphalt film thickness are documented. â¢ 2005âOne level of gyration for NDesign is selected based on the aggregate structure locking point of Georgia mixtures (65 gyrations). chapter five CASE EXAMPLES
77 â¢ Likely to show an accumulation of coarser particles behind the paver screed at the auger gear box (center of screed), at screed extensions, and at the outside edges of the screed (e.g., longitudinal joints); and â¢ Permeable mixtures at locations that hold moisture longer after rain events. GDOT identified two critical factors related to low asphalt contents and low film thicknesses that were evaluated with extensive investigations: â¢ Potential overcompaction of mix design samples [i.e., number of design gyrations (NDesign) too high], and â¢ Overestimation of the contribution of RAP asphalt content to the total effective asphalt content of the mixture. The first extensive GDOT study evaluated the initially selected Superpave levels of compaction (50, 75, 100, and 125 for NDesign) may be overcompacting the mixtures. Overcompacting the mixtures would result in selecting a too low design asphalt content. GDOT explored this pos- sibility by determining the number of gyrations necessary to reach the locking point for a large number of samples and a wide range mixture types. The locking point is when the sample height is constant for three or more consecutive gyrations. Results showed the locking point for GDOT mixtures was consistently between 60 and 68 gyrations for dense-graded mixtures. Based on this study, a single Ndesign of 65 gyrations was selected for the majority of Georgia DOT mixtures. Exceptions to the single gyration level include the GDOTâs 4.75-mm (No. 4) mixtures, which have a locking point of 50 gyrations and SMA mixtures are designed using 35 gyra- tions. GDOT had the National Center for Asphalt Technology (NCAT) verify locking point selections. Georgiaâs Early Experiences with Superpave and RAP Increased evidence of segregationEvidence of moisture intrusion in pavement less than 3 years old Coarse streaking in mixtureDry and quickly aged appearance FIGURE 26 Examples of pavement conditions after initial implementation of Superpave and increasing percentages of RAP (Source: Hines 2015).
78 Between 2005 and 2010, performance testing for mixture approval was added to GDOTâs mix design procedures. The Hamburg wheel tracking device is used to evaluate the mixture rutting potential. Permeability is evaluated using the ASTM PS129-01 Standard Provisional Test Method for Measure- ment of Permeability of Bituminous Paving Mixtures Using a Flexible Wall Permeameter. These changes in NDesign encouraged the use of finer, more uniformly-graded gradations that are less prone to segrega- tion. However, the mixtures still looked dry when using RAP percentages approaching 25%. At this time the entire RAP asphalt was considered to contribute to the total asphalt con- tent of the mixture. That is, the asphalt availability factor for the RAP was 1. In 2012, GDOT conducted a second extensive laboratory study to investigate the possibility that not all RAP asphalt was contributing to the total useful (effective) asphalt con- tent. Because no methodology was, and still is not, standard methodology for determining the RAP asphalt availability factor, GDOT used an approach based on its experience and performance-based testing. The steps used for the laboratory study are: â¢ Step 1: Determine the amount of RAP asphalt that is transferred to virgin aggregate in the plant before the addition of virgin asphalt (dry mixing). â¢ Step 2: Visually estimate the percentage of RAP asphalt remaining on the surface of the RAP particles after dry mixing (effective RAP asphalt). â¢ Step 3: Correct the original optimum asphalt content from the mix design procedure to account for RAP asphalt that is not useful (i.e., asphalt availability factor). â¢ Step 4: Ensure the mixture still meets performance-based mixture testing. Step 1: Transfer of RAP Asphalt to Virgin Aggregate The following methodology was used to visually estimate the likelihood of RAP asphalt transfer to virgin aggregate: â¢ 25% RAP by mass of virgin aggregate was batched and kept at room temperature. â¢ Known mass of light-colored virgin aggregate (No. 6 stone) was preheated at 400Â°F (204Â°C), which was used to approximate superheating the virgin aggregate at the asphalt plant before dry mixing. â¢ Laboratory pugmill mixer was preheated, the super- heated virgin aggregate was added, followed by the room temperature RAP. Materials were mixed for one minute. â¢ Mixture was removed from pugmill, cooled, and the light-colored coarse virgin aggregate particles were separated from the RAP. â¢ Change in the mass of virgin aggregate owing to the transfer of the RAP asphalt was calculated. The results showed only a limited transfer of RAP asphalt was transferred to the virgin aggregate (Figure 27). The RAP asphalt remained on the RAP surface and did not appreciably liquefy and transfer. Based on these results, the RAP asphalt was considered to act more like a partial precoating of the RAP particles rather than an asphalt replacement that can completely and homogeneously blend with the virgin asphalt. Step 2: Estimating Effective RAP Asphalt The second step was to visually estimate the amount of asphalt remaining on the RAP aggregate that acts as a precoating of the RAP aggregate surface. Multiple RAP stockpiles were sampled from around the state and evaluated by the following methodology: â¢ Part 1: â Determine RAP asphalt content using the ignition oven. Pugmill-mixed No. 6 Stone (light color) and RAP Manually separated No. 6 Stone with limited evidence of RAP binder transfer FIGURE 27 Georgia evaluation of potential RAP binder transfer to virgin aggregate during production (Source: Hines 2015).
79 â Mix RAP aggregate remaining at the end of ignition oven testing increasing percentages of virgin asphalt in increments of 0.25% to 0.5%. â¢ Part 2: â Preheat RAP to a temperature achieved during dry mixing at the plant. â¢ Part 3: â Compare coating on the RAP aggregate mixed with various percentages of virgin asphalt (Part 1) to the coating on the preheated RAP (Part 2) (Figure 28). The effective RAP asphalt content was calculated as the ratio of the percentage of virgin asphalt that is to be added to the RAP aggregate so that it appeared similar to the preheated RAP: = ï£«ï£ ï£¶ï£¸ï£« ï£ï£¬ï£¬ ï£¶ ï£¸ï£·ï£· Effective Asphalt Content Ratio Match of AC% of virgin asphalt and RAP aggregate RAP AC% from ignition oven 100 For example, the preheated RAP in Figure 28 had an asphalt content of 4.46% and it took 2.75% of virgin asphalt added to the RAP aggregate (after ignition oven testing) to produce a mixture with a similar appearance: ( )= =Effective Asphalt Content Ratio 2.75%4.46% 100 61.7% The effective asphalt content contribution from the RAP is about 61.7%. After discussions of the results with GDOT contractors, a compromise was reached that assumes an effective asphalt content ratio of 75% (i.e., asphalt availability factor of 0.75 for RAP). Georgia contractors are credited with (paid for) 75% of the asphalt content in their RAP stockpiles. Step 3: Corrected Optimum Asphalt Content The original optimum asphalt content, OOAC, from the initial mix design is still calculated as: ( )( ) = + OOAC % virgin asphalt % RAP % RAP asphalt content The effective RAP asphalt, which is referred to as the cred- ited asphalt content (CAC) to the contractor, is calculated as: ( )( ) ( )=CAC % RAP % RAP asphalt content 0.75 and the noncredited asphalt content (NCAC) is the difference between the RAP asphalt content and the percentage of RAP credited to the contractor (75%): ( ) ( ) ( ) ( ) ( ) = â NCAC % of RAP % RAP asphalt content % RAP % RAP asphalt content 0.75 GDOT increases the original optimum asphalt content adding this percentage of virgin asphalt. This value is the corrected optimum asphalt content: ( ) = + â ï£«ï£ ï£¶ï£¸ï£®ï£°ï£¯ ï£¹ ï£»ï£º COAC OOAC % RAP % RAP asphalt contentCAC virgin asphalt For example, a mix design for a 12.5-mm gradation asphalt mixture with 30% RAP (0.30 in decimal format) and a RAP asphalt content of 5.75% has an OOAC of 5.50%: ( )( )= + =OOAC 5.75% 0.30 3.78% 5.50% The originally determined percentage of RAP asphalt used to calculate the optimum asphalt content is 1.73% and the virgin asphalt content is 3.78%. The percentage of NCAC RAP asphalt content is: ( )( )( )= =NCAC 5.75% 0.30 0.25 0.43% The contractor is credited with a RAP asphalt content of 1.29% and the original percentage of virgin asphalt is increased by 0.43%. The COAC is: = + =COAC 5.50% 0.43% 5.93% Technically, the useful optimum asphalt content is still 5.50% [i.e., 1.29% + 3.78% virgin asphalt + 0.43% (additional) virgin asphalt = 5.50%]; however, the total asphalt content that Heated Original RAP Look of RAP coating after heating Look of RAP aggregate (after ignition oven) with 2.75% virgin asphalt FIGURE 28 Visual comparison of coating of original RAP material with RAP aggregate (Source: Hines 2015).
80 would be measured for the asphalt mixture produced at the plant will be 5.93%. Step 4: Performance and Durability Check Additional samples are prepared using the COAC, and rut- ting potential (APA rut testing) and the moisture sensitivity (durability) are evaluated. As expected, the slight increase in the percentage of virgin asphalt results in a corresponding decrease in air voids that helps improve the mixture durabil- ity. The aggregate structure (gradation) selected during the initial mix design typically still provides the mixture with acceptable rut resistance even with the increases asphalt content. The corrected optimum asphalt content calculation changes were incorporated into the GDOT 2012 specifications and the agency is routinely approving contractor mix designs with 30% RAP. The first high RAP and increased asphalt content surface mixtures were placed in 2012. Mixtures looked well-coated and uniform when placed, and after more than 2 years show no initial evidence of early pave- ment distresses (Figure 29). An additional benefit to GDOT is the reduction in contractor penalties for out of specifica- tion mixtures (Figure 30). CASE EXAMPLE NO. 2: SOUTHEASTERN CONTRACTORâS WORKING WITH GEORGIA DEPARTMENT OF TRANSPORTATION Contractors who typically produce and place high RAP asphalt mixtures for GDOT were asked to complete the same survey that was sent to state construction engineers. These large contractors also have some experience placing RAS and/or combination RAP and RAS mixtures for nonstate agency clients. Five large contractors with multiple plants and contractor laboratories responded to the request for infor- mation. These contractors conduct business in five other Southeastern states (Alabama, Florida, North Carolina, South Carolina, and Tennessee) and provided information about different laboratory practices, various types and ages of asphalt plant types, and placing these mixtures for dif- ferent clients. All six of the contractors indicated that RAP is available in their states, but noted the availability of RAS is limited to one or more state districts or to only local areas within some states (Table 79). Contractors report using various asphalt availability factors for recycled materials (Table 80). Four contractors use a factor of 1 for RAP (i.e., 100% RAP asphalt is useful) and two con- tractors use agency-specified asphalt availability factors. As Surface mixes with more than 25% RAP performing well after 2+ years Improved uniformity in the mix texture and well-coated After Implementation of Corrected Optimum Asphalt Content FIGURE 29 Look of high RAP pavements after implementation of the corrected optimum asphalt content (Source: Hines 2015). FIGURE 30 Impact of RAP mix design changes on contractor disincentives (Source: Hines 2015).
81 the percentage of RAP increases to 25% or more, the clients are more likely to specify the RAP asphalt availability factor. From four to six agencies specify the RAS or combination of RAP/RAS asphalt availability factors. One contractor noted that one of its clients used an ABR of 40% for RAS mixtures. Two contractors have experience with clients using the RBR (recycled binder ratio). South Caro- lina clients typically set a limit on the percentage of recycled asphalt rather than using one of the established ratios for con- trolling the percentage of virgin asphalt in the recycled material asphalt mixture. The contractors provided a wide range of responses for selecting the virgin asphalt grade that reflects the wide range of their clientâs preferences: â¢ Two contractors noted that they have the option for selecting the virgin asphalt grade. â¢ None of the contractors âbumpâ the virgin asphalt grade temperatures. â¢ One contractor bases the virgin asphalt grade selection on the recovered asphalt properties. â¢ One contractor does not make any adjustments. â¢ Four contractors noted that the agency, or other clients, specify the grade of the virgin asphalt. â¢ One contractor indicated that the state agency defines the virgin asphalt grade based on the percentage of RAP in the mixture. â¢ Two contractors set the percentage of recycled materials to be used in the mixture and then select the virgin asphalt grade. â¢ One contractor uses a softening or rejuvenator additive for the stiffer recycled material asphalt, then selects the virgin asphalt grade. â¢ One contractor verifies that the combined mixture asphalt properties meet composite viscosity requirements. Recycled Material Properties Recycled material asphalt content is determined by all of the contractors that use the ignition oven method. The following comments about ignition oven correction factors were provided: â¢ None used (two contractors). â¢ Use a (ignition oven) correction factor on all mixtures by mixing samples at optimum (asphalt content) then TABLE 79 AVAILABILITY OF RECYCLED MATERIALS IN SIX SOUTHEASTERN STATES Supply and Demand: Which types and percentages of recycled materials used in asphalt mixtures can be limited by the available supplies. Also, an overabundance of recycled material(s) can result in various supplyâdemand competitions. Please indicate if recycled materials supplies are available statewide, on a district-by-district basis, or only through a few local material recyclers. Also, indicate if there is any excess of recycled materials (i.e., more supply than demand). Availability of Materials Statewide In One or More Districts/Regions Limited to Local Areas Number of Contractors How widely available is RAP throughout your state? 4 1 0 How widely available is RAS throughout your state? 2 1 0 Is there an excess of shingles (RAP) in your state? 0 2 2 Is there an excess of shingles (RAS) in your state? 0 0 2 Do RAP and RAS compete for use in the tonnage of asphalt mixtures produced in your state? 0 1 2 TABLE 80 CONTRACTOR EXPERIENCE USING ASPHALT AVAILABILITY FACTORS Survey Question: For the purposes of mix designs, indicate which âphilosophyâ is used to establish the contribution of the recycled material asphalt. Materials Number of Contractors 100% Available for Mixture (availability factor = 1) 0% (âBlack Rockâ) (availability factor = 0) Agency-Assumed Percentage of the Total Recycled Asphalt Content 25% or less RAP 4 0 2 More than 25% RAP 3 0 3 RAS, manufacturer waste 2 0 4 RAS, tear-offs 1 1 4 RAS, any combination 0 1 5 RAP and RAS combination 0 0 6 n = 6.
82 burning. The difference between what is burned and the optimum is the (ignition oven) correction factor. â¢ (Ignition) oven correction based on known batched sample of total mixture. â¢ No correction for RAP stockpile burns. The recycled material asphalt is extracted and recovered by five contractors. Only one contractorâs laboratory uses vacuum solvent extraction with Bioact. Recovery methods include Abson (three contractors), Rotavapor (one contractor), or the combination extraction/recovery AASHTO T319 method (one contractor). Four contractors indicated that recycled asphalt is recovered for asphalt testing in some of their laboratories; however, the samples are then submitted to the client for test- ing. Two contractors work with clients that perform their own extractions and recoveries. Once the recycled material asphalt is recovered, the asphalt high temperature shear modulus, G, is determined using the DSR. The G* of the as-recovered asphalt (three contractors) and after RTFO conditioning (one contractor) are the only asphalt properties that are usually evaluated. One contractor uses absolute viscosity testing for some of its clients and back calculates to determine the absolute viscosity for another client. All six of the contractors determine the washed aggre- gate gradations (i.e., sieve analysis, washed sieve for minus 0.075%) for RAP after ignition oven testing and two contrac- tors measure these properties for RAS aggregates (Table 81). Source RAP or RAS aggregate properties are not evaluated, although two contractors mentioned that they look at the aggregate group, class, or petrographic analysis. Mix Design Samples Four of the contractors dry recycled materials prior to batch- ing, use additional sieving of the recycled materials for batching, and heat the virgin aggregate and RAP separately (Table 82). One contractor considers heating prior to mixing sufficient to dry out the recycled materials. How Materials Are Batched for Heating Number of Contractors Drying Before Batching Dry before batching 4 Consider heating for mixing sufficient 1 How Material Is Processed for Batching Batch as stockpiled 1 Additional sieving for tighter gradation control 4 How Material Is Combined, or Not, for Heating Heat aggregate and RAP separately 4 Combine aggregate and RAP before heating 1 Heat aggregate and RAS separately 1 Combine aggregate and RAS before heating 1 Heat combined RAP and RAS separately 1 Combine RAP and RAS before heating 1 TABLE 82 CONTRACTOR PRACTICES FOR DRYING AND COMBINING MATERIALS TABLE 81 RECYCLED AGGREGATE TESTING Survey Question: Indicate which aggregate specification tests are conducted for the recycled material aggregate. (Check all that apply.) Material Gradation Minus 0.075-mm by Washing Flat and Elongated Fractured Faces Fine Aggregate Angularity Sand Equivalent Number of Contractors Ignition Oven RAP, after ignition oven 6 6 2 2 2 2 RAS, after ignition oven 2 2 1 1 1 0 Solvent Extraction RAP, after solvent extraction 0 0 0 0 0 0 RAS, after solvent extraction 1 1 1 1 1 0 Only determine properties for entire mixture with the recycled materials after either solvent or ignition oven testing. 0 0 0 0 0 0
83 Additional comments provided by the contractors included: â¢ Recycled materials are dried before batching: â In air â In an oven at: n 140Â°F (60Â°C) n 220Â°F (140Â°C) n 300Â°F (149Â°C). â RAS is dried at 150Â°F (66Â°C) by one contractor. â¢ Recycled materials are occasionally visually checked for contaminates including extra sand. The order of addition of materials to the mixing bowl is generally consistent between the different contractor labora- tories. Aggregates are added first, followed by the recycled materials, with the asphalt last (Table 83). All mixing is done until the components look uniformly distributed or coated. None of the contractors use a specific time for mixing. When lime is used as an anti-stripping additive, it is added between the aggregate and the RAP. Although the order of addition for mixing is consistent, the temperatures and times used to preheat the materials vary considerably, as indicated by respondentâs comments: Aggregates are heated for: â¢ 4 hours at 230Â°F (110Â°C) â¢ 3 to 4 hours at 390Â°F (199Â°C) â¢ 2 hours at 325Â°F (163Â°C) â¢ 1 hour at 300Â°F (149Â°C) â¢ 2 hours at 375Â°F (191Â°C). RAP is heated for: â¢ 1 hour at 300Â°F (149Â°C) â¢ 2 hours at 325Â°F (163Â°C) â¢ 30 minutes at 300Â°F (149Â°C) â¢ 2 hours at 140Â°F (60Â°C). RAS or a combination of RAP and RAS is heated for: â¢ 2 hours at 325Â°F (163Â°C) â¢ 8 hours at 140Â°F (60Â°C). Short-term aging for mixtures with RAP is accomplished using: â¢ 2 hours at 300Â°F (149Â°C) (three contractors) â¢ 2 hours at 310Â°F (154Â°C) (one contractor). The short-term aging for mixtures with RAS or a combination of RAP and RAS is completed using: â¢ 2 hours at 300Â°F (149Â°C) (one contractor) â¢ 2 hours at 325Â°F (163Â°C) (one contractor). The temperature of the mixtures is measured using a probe in the material while in the oven (two contractors) or immediately after removing from the oven (three contractors). The levels of compaction vary by client and by the type of mixture: â¢ 50 to 75 wear and binder courses (one contractor) â¢ 65 for any mixture type (two contractors) â¢ 50 for SMA (can only use 25% or less RAP) courses (one contractor) â¢ 35 for SMA courses (one contractor). Three of the contractors responding to the survey believe it is more difficult to meet air void requirements when the mix- tures have more than 25% RAP (Table 84). Only one of these contractors believes it is difficult to meet VMA, VFA, and dust-to-asphalt ratio requirements with high RAP mixtures. One contractor believes that any of the volumetric require- ments may be difficult to meet when using a combination of RAP and RAS. Materials Order of Materials Added to Mixing Bowl 1st 2nd 3rd Aggregates, all Fractions 4 0 0 RAP, Coarse 0 3 0 RAP, Fines 0 4 0 RAS 0 1 0 Asphalt 0 0 3 Rejuvenator 0 0 1 TABLE 83 ORDER OF ADDITION OF MATERIALS FOR MIXING Recycled Materials Difficult to Obtain Mix Design Volumetric Properties Air Voids, % VMA VFA Dust-to- Asphalt Ratio 25% or less RAP 0 0 0 0 More than 25% RAP 3 1 1 1 RAS mixtures 0 0 0 0 RAP and RAS combination mixtures 1 1 0 1 TABLE 84 PERCEPTIONS OF INCREASED DIFFICULTY IN OBTAINING VOLUMETRIC PROPERTIES
84 Performance testing of mixtures is limited to using the APA to evaluate the rutting potential during the mix design. A few of the contractors indicated that they do not evaluate rut resis- tance; however, some of their clients do use the Hamburg (dry, wet) or AMPT. None of the contractors or their client labora- tories evaluates stiffness or any type of cracking. This may be because rutting is the primary type of pavement distress of concern in the hot, wet southeastern region of the country. Recycled Material Stockpiling and Processing Collecting, processing, and stockpiling recycled materials can require additional permits, additional storage area, and well- drained stockpiling areas for both RAP and RAS stockpiles. Fugitive dust control is essential during RAP (one contractor) and RAS grinding (two contractors). Certification documents for contaminate-free recycled materials are not necessary; how- ever, one contractor does evaluate the RAP for contaminates. Noise permits are not required for grinding operations. Contractor perspectives about RAP and RAS processing are described here. RAP Processing All of the contractors collect large quantities of RAP before processing, which is usually done in optimum weather condi- tions. Hot and/or wet weather can bog down the crushing pro- cess, blind screens, and reduce the rate of processing. Only one contractor noted any need for time limitations between processing and using RAP. Contractors typically process RAP at the asphalt plant site and sufficient RAP is frequently or occasionally (three contractors) stockpiled at the start of a project to complete the project. Only two contractors noted that they fraction- ate RAP, regardless of the percentage of RAP used in the mixture, by splitting on a single size screen. The coarse RAP is the material retained on the screen and the fine RAP is the material passing the screen. Contractors reported that the screens and sizes used are: â¢ Split on the 4.75-mm (No. 4) sieve. â¢ 19.0-mm to 4.75-mm (Â¾-in. to No. 4) and passing the 4.75-mm (No. 4) sieve. â¢ Minus 12.5-mm (Â½-in.). RAP QC testing is conducted for every 1,000 tons (three contractors). One contractor tests every 500 tons for both RAP and RAS. Asphalt content is determined with the ignition oven and the remaining aggregate is used to determine the gradation. Only one contractor evaluates the fine aggregate angularity for the RAP aggregate. None of the contractors determine the aggregate bulk specific gravity or the bulk specific gravity or theoretical maximum specific gravity of the recycled materials. Four contractors determine the moisture content of the RAP using AASHTO T329 (three contractors) and another method (not defined, one contractor). Contaminates in the RAP stock- piles are evaluated by two contractors. RAS Processing Three contractors indicated that they do not use RAS, and only one contractor rarely uses RAS. However, several of the con- tractors have placed test sections either for their own research or at the request of their clients. Only two contractors have experimented with using a combination of RAP and RAS, but no additional information was provided. Based on their limited previous experience, the contractors have used various maximum RAS grinding sizes (i.e., 100% passing): â¢ 19-mm (Â¾-in.) sieve (two contractors). â¢ 12.5-mm (Â½-in.) sieve (three contractors). â¢ 9.5-mm (3â8-in.) sieve (two contractors). Sufficient RAS was processed by one contractor to com- plete the project prior to the start of construction. Trying to grind RAS in hot, rainy weather caused problems by blinding screens, clumping, and sticking to conveyors. Adding sand to the RAS during or after processing helped keep the RAS from clumping (two contractors) and approximately 1% of the water was used to cool the grinding teeth (one contractor). RAS stockpiles, either unprocessed or processed, are rarely covered. One contractor noted a time delay because of the approval process required for testing the recovered RAS asphalt and only one contractor commented that the contaminates in the RAS stockpiles were measured. Asphalt Mixture Production and Placement Large contractors produce asphalt mixtures in multiple states and have a range of plant types. Each contractor provided information about the plant adjustments and modifications needed to produce high RAP, RAS, and/or a combination of RAP and RAS mixtures for multiple types of plants. Contractors generally believe their current metering meth- ods and sensors are capable of feeding the appropriate amount of recycled materials into the plant. Any type of recycled mate- rial stockpiles can crust, clump, or bridge over the belt weigh scales. One contractor uses in-line crushers to size recycled materials as they are fed into the plant, which helps with break- ing up any clumping. Additional cold feed bins help increase the percentage of RAP or the use of RAS in the mixture. Plant operations occasionally find it necessary to slow the production rates for longer drying times and better mixing when using more than 25% RAP or RAS. Plant temperatures may also have to be raised. Mixtures with more than 25%
85 RAP may need to limit the silo storage time to prevent the mixture from getting too stiff. Batch plant (three contractors) adjustments and/or modi- fications that help improve the amount of recycled material that can be used in the plant include: â¢ Screw conveyor or belt scale moving recycled materials into the pug mill. â¢ Additional venting capability on weigh box to accom- modate steam produced when cold recycled materials are combined with hot aggregate. â¢ Plant configuration for adding recycled material bypasses main vibrating screen and drops directly into the No. 1 bin. â¢ Using a separate unit for drying, proportioning, and feeding recycled materials directly into the pug mill. Parallel flow drum plant (four contractors) adjustments and/or modifications include: â¢ High percentages of RAP (>25%), RAS, or combina- tions of RAP and RAS frequently cause a problem with higher drum exhaust entering the baghouse. â¢ Changes to the fighting in drum to help with heat trans- fer, mixing, and retention time in drum. â¢ Recycled material enters the drum near the center. â¢ Entry collar moved closer to the discharge end of the drum to accommodate higher percentages of recycled materials. Counterblow drum plants (five contractors) for which infor- mation was provided is either a single drum (one contractor) or a double drum (three contractors). Adjustments and/or modifi- cations include: â¢ High percentages of RAP (>25%), RAS, or combinations of RAP and RAS tend to cause a problem with higher drum exhaust entering the baghouse. â¢ Changes to fighting in drum to help with heat transfer, mixing, and retention time in drum. â¢ Improved heat transfer to dry and heat the increased amount of RAP. â¢ Have used warm mix asphalt technology to help reduce exhaust gas temperatures. Mixtures with more than 25% RAP typically flow differ- ently from the haul truck into the paver hopper. One respon- dent described âdifferentlyâ as âmoves in clumps more than it flows.â Kicker paddles help move the stiffer mixtures under the gear box. With uniformity and density at joints, contractors generally think their current metering methods and sensors are capable of feeding the appropriate amount of recycled mate- rials into the plant. Any type of recycled material stockpiles can crust, clump, or bridge over the belt weigh scales. One contractor uses in-line crushers to size recycled materials as they are fed into the plant, which can also be useful in breaking up any clumping. Additional cold feed bins help increase the percentage of RAP or the use of RAS in the mixture. Plant operations occasionally slow the production rates for longer drying times and better mixing when using more than 25% RAP or RAS. Plant temperatures may also need to be raised. Mixtures with more than 25% RAP may find it necessary to limit the silo storage time to prevent the mixture from getting too stiff. Joint density can be more difficult to achieve and visible âlinesâ in the direction of paving can be more noticeable between screed and extensions. Difficulty in moving in a uniform manner tends to make the mixture more likely to segregate. Additional remarks from the contractors included: â¢ High RAP mixtures are stiffer and more temperature sensitive. â¢ RAP asphalt does not transfer or blend; mixtures essen- tially have less film thickness. Nuclear density gauges and cores are used by two con- tractors to monitor mat density and one contractor uses nonnuclear gauges. None of the contractors believe that the recycled material content in the mixture influences any of the gauge readings. One contractor believes that the recycled material content of the mixture may influence the pavement ride quality, whereas two other contractors do not believe this makes any difference to smoothness measurements. Key Points for Field Inspectors Contractors noted that inspectors look for: â¢ Thermal segregation, â¢ Coating of the material, and â¢ Visible contaminates and oversized chunks of RAP. CASE EXAMPLE NO. 3: CONTRACTORâS VIEW OF PRODUCING AND PLACING ASPHALT MIXTURES WITH SHINGLES (MISSOURI) A Missouri contractor presented key issues with designing and producing RAS mixtures for the North Central Asphalt Users and Producers Group (NCAUPG) (Jackson 2012). The major problems identified were: â¢ Contaminates, â¢ Maximum RAS size, â¢ Lift thickness, â¢ Virgin asphalt content, â¢ Virgin asphalt PG grade, â¢ RAS specific gravities, and â¢ RAS moisture content.
86 Contaminates do not mix with the other mixture compo- nents and result in a nonuniform asphalt mixture. Contami- nates can sometimes be identified in the finished pavement surface. RAS mixtures that are placed in lifts that are 25 mm (1 in.) or thinner often show signs of segregation or âshadowing.â Mixtures with too little virgin asphalt can meet the mix design criteria, but still look âdry.â Insufficient virgin asphalt content mixtures are less durable and exhibit early signs of pavement distresses related to insufficient asphalt film thick- ness. Mix design calculations for the amount of new (virgin) asphalt depend on a number of other test results, estimates of other properties from historical records (e.g., ignition oven correction factors for problematic aggregate mineralogy), materials suppliers (e.g., virgin asphalt-specific gravities), and estimates of how much of the recycled material asphalt actually contributes to the total asphalt content. RAS Contaminates It is important that the preprocessed shingles be as free of contaminates as possible (Figure 31). If the asphalt contractor obtained the ground shingles from a recycled material sup- plier, the supplier needs to have a good QC program in place. If the contaminates are not removed prior to processing, then they are ground up along with the shingles and the resulting processed RAS will contain appreciable amounts of deleteri- ous materials that will not likely meet agency specification requirements (Figure 32). Any asphalt that can be contributed by the RAS may be overestimated because larger sizes have lower surface areas. This limits the contact area between the RAS asphalt and virgin asphalt and therefore limits the blending of the two asphalts. The end result is an underasphalted mixture that looks dry behind the paver. Maximum RAS Size Large particles are difficult to uniformly distribute through- out the mixture, clog up going into the drum (Figure 33), and can be sufficiently large so that they are visible in the mixture behind the paver. A smaller maximum RAS par- ticle size (i.e., a finer grind) helps minimize clumping and improve uniform distribution in the asphalt mixtures. Mis- souri DOT reduced the maximum size to 9.5 mm (3â8 in.) to achieve better distribution of the RAS in the mixture, more potential for contributing to the total asphalt content, and reduce the chance of larger RAS particles popping up in the finished pavement surface (Figure 34). Lift Thickness RAS mixtures tend to cool more quickly than conventional mixtures (less thermal mass). Lifts thicker than 25 mm (1 in.) When deleterious materials (contaminates) are not removed from the RAS supply, they are ground up along with the shingles FIGURE 32 Deleterious materials, if not removed before grinding, end up in the RAS supply (Source: Jackson 2012). Shingles collected in 2010 Shingles collected in 2003 FIGURE 31 Clean supply of RAS is needed prior to processing (Source: Jackson 2012).
87 do not cool as quickly as thin lifts. Using a material transfer device helps keep the mixture blended (i.e., limits segrega- tion) and slows heat loss because of the mass of material in the surge bin. Virgin Asphalt Content The three problems that can lead to calculating an optimum asphalt content during the mix design phase that is too low are described here. Trying to Use Too High a Percentage of RAS Contributing to the Total Asphalt Content If the mixture looks dry coming out of the plant, more vir- gin asphalt is required and the asphalt availability factors are to be re-evaluated. Mix design worksheets are to include the asphalt availability factor for reducing the RAS asphalt included in the calculated total asphalt content. Overestimating the Measured RAS Asphalt Content Measuring the RAS asphalt content requires an understanding of the limitations of the test method used to measure the value. For example, mass loss in an ignition oven needs an ignition oven correction factor for the nonasphalt material that burns off. A lower oven temperature or shorter time is typically used when testing RAS. Mix Design Calculations for the Optimum Asphalt Content Are Acceptable, But the Mixtures Look Dry When Produced at the Plant This is a function of the credit given to the RAS asphalt contri- bution and the percentage of virgin asphalt determined in the mix design calculations. The mixture has to perform in the field and the contractorâs crew still needs to get it placed. If it does not look right or is too stiff to place correctly, then the rea- sonableness of the asphalt correction factor used for the mix design should be assessed. Jackson (2012) suggests an inven- tory of RAS mix designs with proven success in both place- ment and performance should be developed. RAS is typically fed into the drum through the RAP chute RAS needs to ï¬ow through the small entrance from the RAP chute into the drum. Problem: Clumping can clog the chute every 7,000 tons of mix. FIGURE 33 Clumps of RAS can be difficult to feed through RAP chute (Source: Jackson 2012). Size used for processing in 2010 Size used for processing in 2005 Scale: Â½-in increments Scale: Â½-in increments FIGURE 34 Reduced shingle size helps with a more uniform distribution of the RAS in the mixture (Source: Jackson 2012).
88 Missouri DOT changed its specifications in 2012 to address previously identified mix design problems for BP-1, BP-2, BP-3, and bituminous base mixtures. The specification changes include: â¢ Lowered design air voids to 3.5%. â¢ Increased requirements for the BP-1 and BP-2 mixtures to 13.5% and 14.0%, respectively. â¢ Reduced design gyrations from 50 to 35; 35 blow Marshall mix design is still acceptable. Virgin Asphalt PG Grade RAS asphalt is very stiff compared with typical paving grade asphalts. RAS asphalt critical upper and lower PG tempera- tures are significantly higher and do not meet agency speci- fications. Although the higher upper critical temperature is useful for improving the rut resistance of an asphalt mixture by increasing the mixture stiffness, the mixture may be too stiff to resist traffic-related cracking. The increased lower critical tem- perature indicates an increased potential for thermal cracking. Also, using RAS with polymer-modified asphalt can also lead to a stiffer, cracking-prone asphalt mixture. The proper selection of the virgin asphalt PG low temperature helps minimize any increased cracking potential. In Missouri, a PG xx-28 offers better low temperature cracking resistance. RAS Aggregate Specific Gravity The RAS aggregate-specific gravity is required for calculat- ing the VMA. More work is necessary to develop procedures and/or practices for determining this RAS material prop- erty. Missouri DOT adjusted its volumetric requirements for BP-2 or surface leveling mixtures. The VMA requirements increased from 13% to 14%. A range of air voids from 3.5% to 4.5% requirement was changed to a single air void con- tent of 3.5%. The field tolerance for the asphalt content was reduced from 0.5% to 0.3%. RAP Moisture Content Too much moisture in the RAS stockpile can cause the RAS to clump, which interferes with uniform feeding of the material into the plant. Clumps of RAS can clog the RAP chute on a drum mix plant that is also used to add the RAS to the mixture. If the RAS is not fully dried during mixing, then the clumps of RAS do not always fully disperse during mixing. Covering the stockpiles (Figure 35) helps reduce RAS moisture contents and a warm mix asphalt additive with a good surfactant may help disperse clumps during mixing. The RAS moisture content is to be removed during the dry- mixing phase of asphalt mixture production. This requires the asphalt plant operator to increase the temperature used to super- heat the virgin aggregate to remove moisture from the recycled material. The increased plant temperatures also help soften the very stiff RAS asphalt, which improves blending with the vir- gin asphalt. However, the plant temperatures are to be kept low enough so that the mixture temperature at the point of discharge meets the agency requirements. These maximum temperature requirements can have a pay factor (disincentive, penalty) for too-hot asphalt mixtures. In Missouri, the maximum mixture temperature is 350Â°F (177Â°C). CASE EXAMPLE NO. 4: LOCATING AND USING COUNTY DATABASES FOR COLLECTING HIGH RAP PERFORMANCE DATA (MINNESOTA) One of the barriers for agencies to increase the percentage of RAP in their mixtures is the lack of performance data. Cur- rently agencies use higher percentages of RAP in asphalt mix- tures that are placed in the lower pavement layers. This makes it difficult to directly link the percentage of RAP to individual pavement distresses that are measured on the pavement sur- face. Information from contractor associations indicate high RAP mixtures are used in surface courses, but not on state agency projects. This case example demonstrates where dis- tress data can be collected for high RAP surface mixtures and used to evaluate pavement performance. The primary distress of concern for MnDOT is low tem- perature cracking (Johnson et al. 2013). A search of county road databases was conducted for projects that had been constructed with 30% or more RAP and one of two virgin asphalt grades (PG xx-34 and PG xx-28). MnDOT requested that the Minnesota county engineers provide information about roadways that had been constructed using RAP and could be accessed using the MnDOT pavement management network. The information to access the pavement condition information was: â¢ County name, â¢ Highway number, â¢ Project limits, FIGURE 35 Cover stockpiles to minimize moisture content (Source: Jackson 2012).
89 â¢ Year constructed, â¢ Design type (wear or nonwear), â¢ Mix design record, â¢ Asphalt performance grade, â¢ Total asphalt content (recycled asphalt plus virgin asphalt), and â¢ Percentage RAP. The search of the Minnesota databases provided a collec- tion of project information that was used to link pavement performance to the virgin asphalt grade, design asphalt con- tent, percentage of RAP, age of the roadway, and the number of projects for each group of variables (Table 85). Projects with no RAP content were also identified that were used as control sections for the pavement performance analysis. The Minnesota County Highway Testing Program was used to further locate specific roadway segment informa- tion to access the pavement performance information in the Pavement Management database. This information included county name, highway number, project limits, survey year, distance, transverse crack count, and other observations (not defined). Once this information was assembled, the county highway performance was developed from a combination of video-log reviews and field inspections (Table 86). The results from this effort produced a good database of high RAP mixtures in wear courses that can be used for con- tinued monitoring of the cracking potential of these mixtures in a cold climate. The analysis of the data allowed MnDOT to identify the percentage of the virgin asphalt in the mixtures as a key factor in the pavement performance. MnDOT speci- fications now limit the minimum percentage of virgin asphalt in the total asphalt content. In this example, nonstate agency projects were used to provide performance information for adjustments to state agency project specifications. CASE EXAMPLE NO. 5: INVESTIGATING TRANSFER OF RECYCLED MATERIAL ASPHALT DURING DRY MIXING Recent research projects have evaluated the amount of recycled material asphalt that can be transferred to the virgin aggregate during the dry mixing time before the virgin asphalt is added. The objectives of these studies were to: â¢ Calculate an approximate amount of RAP asphalt that is available to blend with the virgin asphalt (Georgia study; Hines 2015). â¢ Find out how much of the RAP asphalt is blended with the virgin asphalt under normal (i.e., plant) mixing conditions (Tennessee study; Huang et al. 2005). â¢ Explore how the mixing temperature can soften RAS asphalt so that it can coat the virgin aggregate (Texas study; Zhou et al. 2013). â¢ Investigate the activation (i.e., transfer) of RAP asphalt to virgin aggregate during dry mixing at a batch plant and compare with the transfer obtained with laboratory (dry) mixing (Minnesota study; Johnson et al. 2013). Georgia RAP Study The GDOT RAP transfer study was a part of laboratory studies used to modify GDOT specifications presented in Case Exam- ple No. 1 and will only be summarized here for comparison to other recycled material asphalt dry mixing transfer studies. Light-colored No. 6 stone [25-mm (1-in.) to 4.75-mm (No. 4) sieve sizes] was used so that the finer RAP could be separated from the virgin aggregate after dry mixing. The virgin aggregate was preheated to 400Â°F (204Â°C) and RAP was kept at room temperature to simulate the material temperatures as they are added to the asphalt plant. Both materials were added to a preheated laboratory pugmill and dry mixed for 1 minute. After dry mixing, the virgin aggregate was visually separated into one of two groups: uncoated and partially coated. The per- centage of the aggregate in each group is measured based on the change in weight (mass) of the virgin aggregate. The results showed that only a limited amount of RAP asphalt was trans- ferred to the virgin aggregate (Figure 36). Tennessee RAP Study Researchers evaluated the amount of RAP asphalt that was transferred to virgin aggregate during dry mixing (Huang et al. 2005). Fine RAP [minus 4.75 mm (No. 4)] was dry mixed with various percentages of coarse virgin aggregate (10%, 20%, and 30%). The virgin aggregates were preheated to 374Â°F (190Â°C) and the RAP was kept at ambient temperature. The results showed that only about 11% of the RAP asphalt was transferred to the virgin aggregate. Virgin Asphalt PG Design Asphalt Content, % Virgin Asphalt Content, % % RAP Age, Years No. of Projects 58-28 4.8 to 6.3 3.0 to 6.3 0 to 40 1 to 11 22 52-34 5.2 to 6.1 3.0 to 6.1 0 to 40 3 to 11 39 58-34 5.5 to 6.2 4.3 to 6.2 0 to 20 1 to 5 6 64-28 6.2 6.2 0 8 1 Source: Johnson et al. (2013). Mix design data were used for asphalt content information. Results may change if using production data. % RAP information includes 37 high-RAP data points (30% or more RAP content). TABLE 85 SUMMARY OF DATA COLLECTED FROM COUNTY ENGINEERS
90 Year Type (lift in.) PG Grade Lift Thickness RAP, % Total Asphalt Content Virgin Asphalt Added No. of Cracks Length, miles ABR Cracks per Mile PG 52-34 Data (control sections) 2009 Wear (1.5), 1 52-34 1.5 0 6.1 6.1 80 2.059 1.00 38.9 2009 Wear (1.5), 1 52-34 1.5 0 5.9 5.9 170 4.999 1.00 34.0 2009 Wear (1.5), 1 52-34 1.5 0 6.3 6.3 56 1.120 1.00 50.0 PG 52-34 Data (30% RAP) 2009 Wear (0.5), 2 52-34 0.5 30 5.4 4.0 1 0.303 0.74 3.3 2009 Wear (0.5), 2 52-34 0.5 30 5.4 4.0 1 0.037 0.74 27.0 2009 Wear (0.5), 2 52-34 0.5 30 5.4 4.0 25 0.848 0.74 29.5 2009 Wear (0.5), 2 52-34 0.5 30 5.4 4.0 14 1.019 0.74 13.7 2009 Wear (0.5), 2 52-34 0.5 30 5.4 4.0 22 0.199 0.74 110.6 2009 Wear (1.5), 1 52-34 1.5 30 5.1 3.8 14 1.019 0.75 13.7 2009 Wear (1.5), 1 52-34 1.5 30 5.1 3.8 3 0.040 0.75 75.0 2009 Wear (1.5), 1 52-34 1.5 30 5.1 3.8 1 0.303 0.75 3.3 2009 Wear (1.5), 1 52-34 1.5 30 5.1 3.8 9 0.381 0.75 23.6 2009 Wear (1.5), 1 52-34 1.5 30 5.1 3.8 1 0.037 0.75 27.0 2009 Wear (1.5), 1 52-34 1.5 30 5.1 3.8 25 0.848 0.75 29.5 2006 Wear (1.5), 1 52-34 1.5 30 5.3 3.6 130 5.100 0.68 25.5 2009 Wear (1.5), 1 52-34 1.5 30 5.3 3.6 1 0.044 0.68 22.7 PG 58-28 Data (control sections) 1999 Wear (1.5) 58-28 1.5 0 6.1 6.1 410 4.872 1.00 84.2 2003 Wear (1.5) 58-28 1.5 0 6.1 6.1 766 3.196 1.00 239.7 2009 Wear (1.5), 1 58-28 1.5 0 5.8 5.8 14 1.510 1.00 9.3 2007 Wear (1.5), 1 58-28 1.5 0 6.1 6.1 14 1.510 1.00 9.3 PG 58-28 Data (30% and 40% RAP) 2003 Wear (1.5) 58-28 1.5 30 5.3 3.6 51 1.837 0.68 27.8 Wear 58-28 1.5 30 5.3 3.6 109 2.727 0.68 40.0 (1.5), 1 2007 Wear (1.5), 1 58-28 1.5 30 5.3 3.6 88 2.765 0.68 31.8 2007 Wear (1.5), 1 58-28 1.5 30 5.3 3.6 225 8.163 0.68 27.6 2009 Wear (2.5), 2 58-28 2.5 40 5.2 3.0 51 1.837 0.58 27.8 2009 Wear (3.0), 1 58-28 3 40 5.2 3.0 109 2.727 0.58 40.0 2009 Wear (3.0), 3 58-28 3 40 5.2 3.0 88 2.765 0.58 31.8 2005 Wearing (1.5) 58-28 1.5 40 5.2 3.0 225 8.163 0.58 27.6 Source: After Johnson et al. (2013). 2007 TABLE 86 EXAMPLE OF AVAILABLE PERFORMANCE DATA FOR SURFACE MIXTURES FOR MINNESOTA COUNTY ROADWAYS
91 FIGURE 36 Appearance of No. 6 limestone after dry mixing with RAP in laboratory pugmill (Source: Hines 2015). FIGURE 37 Example of RAP asphalt transfer to virgin aggregate after 3 minutes of laboratory mixing at 190Â°C (Source: Huang et al. 2005). Texas RAS Study Researchers at the Texas A&M University Texas Transportation Institute in cooperation with TxDOT and FHWA conducted a study to characterize and identify the most effective uses of RAS in asphalt mixtures (Zhou et al. 2013). A component of this research was an evaluation of the plant production tempera- tures required for the RAS asphalt to transfer to the virgin aggre- gate during dry mixing. The virgin white limestone aggregate was dry mixed with each of two types of RAS (manufacturer waste and tear-offs) at one of four temperatures [143Â°C, 149Â°C, 163Â°C, and 200Â°C (290Â°F, 300Â°F, 325Â°F, and 392Â°F)] using a batching ratio of 80% virgin aggregate to 20% RAS (Figure 37). Mixing of the two materials was accomplished by: Manufacturing Tear Oï¬s RAS Binder Transfer Study 290Â°F (143Â°C) 300Â°F (149Â°C) Typical Texas mixing temperature 325Â°F (163Â°C) 392Â°F (200Â°C) FIGURE 38 RAS asphalt transfer to virgin aggregate over a range of temperatures (Source: Zhou et al. 2013). â¢ Screening the virgin aggregate to obtain the material passing the 12.5-mm (Â½-in.) sieve and retained on the 9.5-mm (3â8-in.) sieve, which was then washed, dried, and heated overnight at mixing at the test temperature. â¢ Heating the RAS overnight at 60Â°C (140Â°F). â¢ Manually mixing the virgin aggregate and RAS followed by short-term aging of the mixed materials at the test temperature. â¢ Mixing the short-term aged blend of virgin aggregate and RAS in a bucket mixer for 2 to 3 minutes. â¢ Short-term aging of the blend again at the test tempera- ture for another 2 hours. â¢ Sieving the virgin aggregate and RAS blend over a 9.5-mm (3â8-in.) sieve. â¢ Visually evaluating the virgin aggregate that is retained on the 9.5-mm (3â8-in.) sieve to estimate the percentage of RAS asphalt transfer (Figure 38).
92 Temperatures and % RAP 420Â°F 10% RAP 490Â°F 24% RAP 400Â°F 24% RAP Samples from haul trucks for Run 1, Run 2, and Run 3 FIGURE 39 RAP asphalt transferred to virgin aggregate after dry mixing for 30 seconds in batch plant (Source: Johnson et al. 2013). Run No. Plant Temp, Â°F RAP Content, % Dwell Time, Seconds Sample Temp., Â°F 1 420 10 30 320 (front of haul truck) 344 (back of haul truck) 2 490 24 30 290 to 300 3 (1st half) 400 24 30 230 (front of haul truck) 3 (2nd half) 375 24 30 225 (back of haul truck) Source: Johnson et al. (2013). TABLE 87 VARIABLES FOR DRY MIXING VIRGIN AGGREGATE AND RAP AT THE BATCH PLANT â¢ Recycled asphalt and aggregate fines formed balls in all of the dry-mixed materials (Figure 40). Four dry-mixed batches were produced in the laboratory using the batch plant temperatures, time allowed for preheating the RAP, and the mixing (dwell) time (Table 88). Most of the laboratory mixture batches were approxi- mately 2,500 grams and prepared in a bucket mixer. The normal preheating temperature used by MnDOT with the bucket mixer is 290Â°F (143Â°C) and the standard mixing time is 10 minutes. The upper temperature for the labora- tory study was limited by the practical operating range of the laboratory oven, which was 320Â°F (160Â°C). The major- ity of the laboratory studies used one of two temperatures, four RAP preheating times, and two mixing times in the laboratory pugmill (Table 89). A limited number of larger batches (15,000 grams) was produced at 300Â°F (149Â°C) using 23% RAP and 50% RAP. Once the virgin aggregate and RAP were dry mixed, the material was manually separated into three groups: uncoated, partially coated, and coated (Figure 41). The percentage of material in each group was determined and the results used The visual evaluations showed the manufacturer waste RAS transferred more asphalt to the virgin aggregate than did the tear-off RAS and most transfer was obtained at the highest temperature of 200Â°C (392Â°F). Although the study showed that RAS asphalt may become sufficiently soft to blend with the virgin asphalt, the high temperature necessary to achieve the best blending (transfer) and the extended time needed for the RAS asphalt to soften enough to transfer were not reasonable conditions for the actual production of asphalt mixtures. Minnesota RAP Study MnDOT conducted a study to assess the transfer of the RAP binder during dry mixing using a batch plant and in a labora- tory setting. The recycled asphalt transfer was evaluated using a modified AASHTO T195-67 Standard Method of Test for Determining the Degree of Particle Coating of Bituminousâ Aggregate Mixtures. The plant was a three-tiered batch plant equipped with six cold feed bins and one RAP belt feed bin. The mixing unit was a twin pugmill type with at most a 0.75-in. clear- ance from the walls and timer controls for wet and dry mix- ing. Plant temperatures and the percentage of RAP added to the pugmill with the virgin aggregate varied (Table 87). The temperature of the aggregateâRAP mixtures was measured at the point of discharge using the integrated plant sensor and a hand-held thermometer. Temperatures were also mea- sured when the materials were loaded into the haul truck. The aggregateâRAP dry-mixed material in the haul trucks was sampled and retained for comparisons of laboratory- produced, dry-mixed materials. Visual observations of the dry-mixed material sampled from the haul trucks showed: â¢ More RAP asphalt transfer (Figure 39) was achieved with the higher RAP content (24%) and at higher temperatures. â¢ RAP asphalt was uniformly transferred to the virgin aggregate at all of the dry-mixing temperatures.
93 Run 1 Run 2 Run 3 FIGURE 40 Balls of asphalt and fines after dry mixing in batch plant (no virgin asphalt) (Source: Johnson et al. 2013). Plant Run No. RAP, % Aggregate Temperature, oF (oC) Time Used to Heat RAP, Minutes Mixing Time, Minutes Completely Coated Particles, %* Partially Coated Particles, %* Uncoated Particles, Particles, %* 1 10% 420 (215) 0.5 0.5 67 33 0 2 23% 490 (254) 0.5 0.5 48 52 0 2, washed 23% 490 (254) 0.5 0.5 53 47 0 3 23% 400 (204) 0.5 0.5 44 56 0 Source: Johnson et al. (2013). *Estimated values from source figure. TABLE 88 LABORATORY DRY MIXING STUDY USING BATCH PLANT VARIABLES Temperature, oF (oC) Time Used to Heat RAP, Minutes Mixing Time, Minutes 290 (143) 1, 90 10 180 1, 5 320 (160) 10, 20, 180, 190 10 Source: After Johnson et al. (2013). TABLE 89 TEMPERATURE AND TIMES USED IN MINNESOTA LABORATORY STUDY 23% RAP, Preheated for 100 min., 300Â°F, Mixed 3 min. 50% RAP, Preheated for 100 min., 300Â°F, Mixed 3 min. 23% RAP, No Preheating. Mixed 3 min. FIGURE 41 RAP transfer to virgin aggregate in laboratory pugmill (Source: Johnson et al. 2013).
94 to calculate the percentage of uncoated, partially coated, and completely coated particles. Visual evaluations showed: â¢ Duplication of dry mixing at the batch plant was not replicated in the laboratory. â¢ No clumping or balling of fines was seen in the laboratory- prepared, dry-mixed batches. â¢ Partially coated aggregates in the laboratory study showed signs of abrasion with little transfer of RAP asphalt. â¢ Large percentages of uncoated particles were seen in all laboratory dry-mixed blends. â¢ The 10% RAP mixtures tended to have higher percent- ages of partially coated, but nearly 0% of fully coated particles (laboratory study; smaller batches). The percentages of the three levels of particle coating were used for various statistical analyses (Pearsonâs correlation coefficients, multi-variable regression equations) to determine which variables had the most significant impact on the transfer of the RAP asphalt to the virgin aggregate by dry mixing in the laboratory. The statistical analysis showed the: â¢ Complete coating modelâMost strongly dependent on the total aggregate retained on the 3â8-in. (9.5-mm) sieve and the percentage RAP. â¢ Partial coating modelâMost strongly dependent on the total aggregate retained on the 3â8-in. (9.5-mm) sieve, mixing time, and the heating time of the RAP. â¢ No coating modelâMost strongly dependent on the percentage RAP and the temperature of the aggregate. The key finding was that a significant amount of recy- cled material asphalt is uniformly transferred to the virgin aggregate during dry mixing at the asphalt plant; how- ever, this transfer cannot be replicated with dry mixing in the laboratory.