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.
The design of PFCs is similar to the design of typical dense- graded HMA mixtures in that the design of PFCs involve four primary steps. First step: select appropriate materials. Materials needing selection include coarse aggregates, fine aggregates, asphalt binder, and stabilizing additives. Second step: blend the selected aggregates to develop a design gradation. Third step: select the optimum asphalt binder content. Fourth step: subject the mixture to performance testing. This chapter presents the current state of practice on the design of PFCs. Sections within this chapter are divided into the four steps in designing PFCs as described above. Materials Selection Materials needing selection include coarse aggregates, fine aggregates, mineral fillers, asphalt binders, and stabilizing additives. The current state of practice for selection of these materials is discussed here. Aggregate Characteristics The survey of agencies described in Chapter 2 included a request for the respondents to rank various aggregate char- acteristics for use in PFCs. Aggregate characteristics included within the survey were abrasion resistance, durability, pol- ish resistance, angularity, shape, cleanliness, and absorption. Results from the survey are illustrated in Figure 13. Respon- dents were requested to rank the various aggregate charac- teristics on a scale of 1 to 7, with 1 being the most important property and 7 being the least important. Based upon the results of the survey, there appears to be three levels of importance. Polish resistance and durability were the most important prop- erties as both of these had the lowest average rankings. The next level of importance includes angularity, abrasion resis- tance, particle shape, and cleanliness. All four of these charac- teristics have reasonably similar average ratings. The final level of importance was aggregate absorption. The average rating of this characteristic was much higher than the other six char- acteristics. These results would indicate that test methods and criterion are needed for abrasion resistance, durability, polish resistance, angularity, shape and cleanliness. Results from the survey agree with the literature reviewed as part of NCHRP 9-41. Desirable aggregate characteristics found in the literature include abrasion/degradation resis- tance, polish resistance, angularity/surface texture, cleanli- ness, particle shape, and durability. Unfortunately, the litera- ture was limited to authors only providing specification values for the different aggregate characteristics. No references were found that provided quantitative evaluations of various levels of different aggregate characteristics in order to optimize the desirable properties for aggregates used in PFCs. As highlighted in the survey results, durability and polish resistance were the two most important aggregate charac- teristics based upon the survey. In Europe, polish resistance also is considered one of the most important aggregate characteristics (21). The polish stone value is the most com- mon requirement specified for ensuring polish resistance (16, 21). The predominant test used to evaluate the durability of aggregates is sulfate soundness. Georgia has a maximum loss of 15 percent when determined using magnesium sulfate (11). Oregon utilizes a maximum loss of 12 percent when using magnesium sulfate (7). The Los Angeles abrasion test is the most common test to evaluate aggregate abrasion/degradation resistance. It is specified both in the United States and internationally. Maxi- mum loss values encountered in the literature ranged from a low requirement of 12 percent (35) to a high of 50 percent loss (11). Within the United States, current recommenda- tions are generally a maximum Los Angeles abrasion loss of 30 percent (4). Coarse aggregate angularity is most often specified as a min- imum number of fractured faces. Most commonly, specifi- cations are for the percentage of aggregates with two or more C H A P T E R 5 Materials and Mix Design 32
fractured faces. Specification values range from a low of 90 per- cent of the coarse aggregates with two or more fractured faces (36) to a high of 100 percent (7, 21). Only one reference was encountered that listed any speci- fication values for fine aggregates. Generally, most references simply stated that the fine aggregate fraction should be crushed indicating an angular material. Kandhal (4) recommended using the uncompacted void content of fine aggregate with a specification minimum of 45 percent. Two tests are generally utilized to specify the desired shape of coarse aggregates, the flakiness index and the flat and elon- gated test. The flakiness index is generally specified in Europe with a maximum requirement of 25 percent (7, 37). Arizona also has utilized this specification for the flakiness index (7). Within the United States, the flat and elongated test is the most common test to define coarse aggregate particle shape. Requirements for flat and elongated tests are generally based upon a ratio of 5:1 (11) though some guidance specifies ratios of 5:1 and 3:1 (4). When a 5:1 ratio is specified, a maximum per- centage of flat and elongated particles requirement of 10 percent is common (11, 36), though some specify a maximum of 5 per- cent (4). When a 3:1 ratio is specified, a maximum requirement of 20 percent is used (4, 36). Aggregate cleanliness is most often specified based upon the sand equivalent test. Specification values for the sand equiva- lent test range from a low of 45 (5, 34) to a high of 55 (7). Asphalt Binders A wide range of asphalt binders have been used in PFC mixes. Both unmodified and modified asphalt binders have been used with success. In the NCHRP synthesis by Huber (7), he reported many different types of asphalt binders. These binders were graded in accordance with the Superpave per- formance grading (PG) system, viscosity grading procedure, and penetration grading system. Within Europe, the asphalt binders were predominantly graded using the penetration grading system. Huber (7) reported on material requirements from Great Britain, Spain, Italy, and South Africa. At the time, Great Britain utilized a 100 pen asphalt binder with and with- out polymer modification. Spain utilized either a 60/70 or an 80/100 pen asphalt binder with polymer modification. Italy also used an 80/100 pen asphalt binder with polymer modifi- cation. Each of these three countries specify either a styrene butadiene styrene (SBS) or ethylene vinyl acetate (EVA) when using polymer modification. Huber (7) indicates that South Africa allows both polymer modification and modification with rubber. Within the United States, Huber (7) reported a wide range of asphalt binders being used. Both PG and viscosity-graded binders were reported. Some U.S. agencies utilized unmodified asphalt binders. For instance, Arizona was specifying a PG 64-16 and Georgia was specifying a PG 67-22 for some OGFC mixes. 33 Survey Rankings for Aggregate Characteristics 0 1 2 3 4 5 6 7 Polish Resistant Durability Angularity Abrasion Resistance Particle Shape Cleanliness Absorption Ranking tnatropmItsaeLtnatropmItsoM Figure 13. Ranking of aggregate characteristics from agency survey.
When utilizing PFCs, most agencies specified modified as- phalt binders. For instance, Oregon specified a viscosity graded AC-30 with 12 percent rubber added to the asphalt binder. Georgia also specified polymer modified binders (7). Alvarez et al. (35) also provided a synthesis on mix design criteria for PFCs. This work was published in 2006, six years after Huberâs synthesis (7). Alvarez et al. (35) also report that asphalt binders used in PFCs are generally modified. Within Europe, polymers also are generally used to modify the asphalt binders. Similar to Huber (7), the polymer types most often cited were SBS and EVA. Great Britain does allow the use of styrene butadiene rubber (SBR) for polymer modification. The only European country not reporting the use of modified asphalt binders was Switzerland. Australia also allows the use of unmodified binders for lower traffic roadways (19, 35). For higher traffic roadways, SBS, SBR, EVA and rubber modified binders are allowed in Australia. Within the United States, Alvarez et al. (35) reported on mix design methods that allowed both polymer modified (type not given) and rubber modified asphalt binders. Proper selection of the asphalt binder to be used within PFCs should be based upon a number of factors. Ruiz et al. (16) state that selection of the asphalt binder should be based upon the weather at the project site and the anticipated traffic vol- ume the roadway will carry. Kandhal (4) also provides similar factors for selection of asphalt binders for PFCs. Generally, the literature indicates that binders with a high stiffness are needed for PFCs, hence most agencies require modified asphalt binders. High stiffness binders are needed to help prevent draindown which promotes thick films of asphalt binder coating the aggregates. Molenaar and Molenaar (38) indicated that stiff, polymer-modified binders also help pre- vent short-term raveling. Short-term raveling was defined as raveling caused by intense shearing forces at the tire/ pavement interface that occurs within newly placed porous asphalt. Ruiz et al. (16) state that asphalt binders that are too soft may tend to bleed during hot weather and lead to rutting problems. Even though stiff binders are desirable, Ruiz et al. (16) also suggest that binders that are too stiff can be detrimental. Asphalt binders that are too stiff may reach a critical hardness earlier which could lead to long-term raveling problems. Stabilizing Additives According to the survey and literature, one of the primary concerns with open-graded mixes is draindown during con- struction. Open-graded mixes have an open aggregate grading with a relatively low percentage of material passing the No. 200 (0.075 mm) sieve. Because of the open grading, the surface area of the aggregate blend is much lower than typical dense- graded mixes, and the low aggregate surface area results in relatively thick asphalt binder films coating the aggregates. According to Watson et al. (39, 40), typical asphalt binder film thicknesses for PFCs are approximately 30 microns compared to approximately 8 microns for dense-graded HMA. At typical production/construction temperatures, the thick film of asphalt binder common to PFCs has a propensity to drain from the aggregate structure, termed draindown (7). In order to reduce the potential for draindown, stabilizing additives are generally incorporated into PFCs. Two types of stabilizing additives can generally be utilized within PFCs: fibers and asphalt binder modifiers. Many different types of fibers have been used within PFCs including mineral, cellulose, asbestos, polypropylene, polyacrylonitrile, glass, and acrylic fibers. According to the results of the agency survey, 85 per- cent of the responding agencies specify the use of fibers within open-graded mixes. This value is significantly higher than the 19 percent of agencies reporting the use of fibers within OGFC mixes in the 1998 survey by Kandhal and Mallick (8). The increase in the percentage of agencies specifying the use of fibers within open-graded mixes is likely an indication of the effectiveness of fibers in reducing draindown potential. Figure 14 illustrates the effect of fiber addition on draindown potential. Data used to create Figure 14 is from research con- ducted by the National Center for Asphalt Technology on PFCs and was published by Watson et al. (40) in a slightly different form. Figure 14 clearly illustrates that the addition of fiber sig- nificantly reduces draindown potential. Similarly, other re- search projects have shown that the use of fibers significantly reduces the potential for draindown (36). According to Pasetto (41), additional benefits can be realized from the addition of fibers within PFC mixes. Pasetto (41) showed that the addition of fibers increased the strength of PFC mixes as measured by Marshall stability and indirect tensile testing. Additionally, the use of fibers improved the durability of PFC mixes as measured by the Cantabro Abrasion test. As stated previously, a wide range of fiber types have been used in open-graded mixes. Within the United States, the most common fiber types used are cellulose and mineral fibers. These two fiber types also are common in Europe (21) and Australia (19). Addition of fibers is generally at a dosage rate between 0.1 and 0.5 percent, by total mix mass. An important point made by Decoene (18) is that the selected fibers must be resistant to temperatures above typical production temperatures. This is especially true when using organic fibers. The other type of stabilizing additive commonly used in open-graded mixes is asphalt binder modifiers. These mod- ifiers are generally polymers or rubber particles. With respect to draindown, these modifiers serve to increase the viscosity (stiffness) of the asphalt binder, helping to maintain the asphalt binder within the aggregate structure. The benefits of modi- fied asphalt binders are not limited to helping prevent drain- down. A series of reports and papers from the National Center 34
for Asphalt Technology (40, 42, 43) have shown that the use of modified asphalt binders that provide higher stiffness at typical in-service temperatures help provide increased dura- bility in the laboratory. These results match field experiences described by Huber (7) and the results of the 1998 survey of U.S. agencies by Kandhal and Mallick (8). Huber (7) indicated that in the past, thick films of unmodified asphalt binder tended to drain downward during hot summer weather due to gravitational forces. The remaining thin films of asphalt binder coating the aggregates would age more rapidly becom- ing brittle, which resulted in raveling. Use of modified asphalt binders helps to retain the thick asphalt binder film, thus improving durability. In addition, research has shown that the use of modified asphalt binders improves the short-term per- formance of PFCs. The increased stiffness of the asphalt binder reduces the potential for traffic dislodging aggregate particles shortly after construction. This early age dislodging of aggregate particles has been termed short-term raveling (38). Fillers/Adhesion Agents A number of agencies from around the world specify the use of fillers or other adhesion agents to improve the bond between aggregates and the asphalt binder. Van Der Zwan et al. (17) state that limestone filler is added during the pro- duction process to improve bonding in the Netherlands. The limestone filler must have a hydrated lime content of at least 25 percent. Australia also requires the addition of a filler to PFC mixes (19). Hydrated lime is the preferred type of filler in Australia; however, portland cement and ground limestone also are allowed. Similarly, Watson et al. (11) stated that hydrated lime is required in PFC mixes in Georgia as an anti-stripping agent. In their 1998 survey of U.S. agencies, Kandhal and Mallick (8) evaluated the reported performance of open-graded mixes with various mix design practices. One of the mix design items included within the evaluation was whether the agency specified fillers/adhesion agents. To better evaluate the infor- mation, Kandhal and Mallick (8) divided the various agen- cies by the Strategic Highway Research Program climatic zones in which each resided. These climatic zones included wet-freeze, wet-no freeze, dry-freeze, and dry-no freeze. Collectively, of the 19 agencies reporting good performance, 53 percent added some type of filler/adhesion agent whether the material was hydrated lime or liquid antistrip materials. Conversely, only 21 percent of the agencies reporting bad performance with open-graded mixes specified the use of fillers/adhesion agents. Interestingly, all of the agencies re- porting good performance within the dry-freeze climatic zone specified hydrated lime, while 75 percent of the agencies reporting bad performance in this climatic zone did not specify fillers/adhesion agents. 35 Effect of Fiber on Draindown Potential 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 Draindown Values Without Fiber, % Dr ai n do w n Va lu es W ith F ib er , % PG 67-22 PG 76-22 Line of Equality Figure 14. Effect of fibers on the draindown potential of PFCs (40).
Selection of Design Gradation The next step within the design of PFC mixes is to utilize the selected aggregates to develop a design gradation (or design aggregate structure). Within a typical mix design, this step may include developing several trial gradations and using mix design criteria to select the most appropriate of the trial grada- tions. Within this section, only typical PFC gradations will be discussed because the following section will provide the differ- ent mix design criteria. The literature review and survey of agencies resulted in a wide range of gradations encountered for PFC mixes. As stated in Chapter 2, nine of the responding agencies categorized their open-graded mixes as PFCs. From these nine agencies, three different maximum aggregate size gradations are specified: 1 in., 3â4 in., and 1â2 in. (25 mm, 19 mm, and 12.5 mm). Here, the term maximum aggregate size indicates the finest sieve which has 100 percent of the aggregates passing. The majority of the gradation requirements encountered throughout the world could potentially be characterized by multiple nominal max- imum aggregate sizes (as defined in Superpave) depending upon the actual blended gradation. Therefore, within this doc- ument gradations only will be discussed by the maximum aggregate size. Oregon was the only U.S. agency that had gradation criteria for a 1 in. (25 mm) maximum aggregate size gradation. This gradation is illustrated in Figure 15. This figure shows that the PFC gradation is gapped on the No. 4 sieve. The allowable filler content for this gradation band is 1 to 6 percent. A number of agencies provide gradation requirements for a 3â4 in. (19 mm) maximum aggregate size. Figure 16 illustrates the various PFC gradation bands specified in the United States. Gradations shown within this figure typically are gapped on the No. 4 sieve; however, some allow for gapping the gradation on the No. 8 sieve. Allowable filler contents range from a low of 1 percent to a high of 5 percent. Interestingly, several agen- cies have identical (or almost identical) gradation requirements for 3â4 in. maximum aggregate size PFCs. Alabama, Georgia, Louisiana, and South Carolina all specify essentially the same gradation requirements. These specifications can be traced to the original Porous European Mix utilized by Georgia (11) and the research conducted by the National Center for Asphalt Technology (40). Louisiana was the sole agency that provided gradation re- quirements for a 1â2 in. (12.5 mm) maximum aggregate size PFC, illustrated in Figure 17. For this gradation band, the aggregates are gapped on the No. 8 sieve. Filler criteria include a minimum of 2 percent and a maximum of 4 percent. From an international standpoint, there are a number of agencies that specify open-graded mixes that would meet the requirements of a PFC. The primary mix design item that was utilized to determine whether the various open-graded mixes would qualify as PFC was air void content at design. In the literature review, open-graded mixes that were specified to 36 0 10 20 30 40 50 60 70 80 90 100 Sieve Size Pe rc en t P as si ng Oregon ¾ in.½ in.3â8 in.No. 4No. 8No. 200 Upper Gradation Limits shown as solid lines, Lower Limits shown as dashed lines. 1 in. Permeable Friction Course Gradations ~ US Agencies Figure 15. 1 in. PFC gradation requirements from U.S. agencies.
37 0 10 20 30 40 50 60 70 80 90 100 Sieve Size Pe rc en t P as si ng Alabama Georgia Georgia 2 Louisiana 2 North Carolina Oregon 2 South Carolina Tennessee ¾ in.½ in.3â8 in.No. 4No. 8No. 200 Upper Gradation Limits shown as solid lines, Lower Limits shown as dashed lines. 1 in. Permeable Friction Course Gradations ~ US Agencies Figure 16. 3â4 in. PFC gradation requirements from U.S. agencies. 0 10 20 30 40 50 60 70 80 90 100 Sieve Size Pe rc en t P as si ng Louisiana ¾ in.½ in.3â8 in.No. 4No. 8No. 200 Upper Gradation Limits shown as solid lines, Lower Limits shown as dashed lines. 1 in. Permeable Friction Course Gradations ~ US Agencies Figure 17. 1â2 in. PFC gradation requirements from U.S. agencies.
have a minimum of 18 percent air voids were considered PFCs. Similar to the survey of agencies described earlier, three differ- ent maximum aggregate size gradations were found in the literature: 1 in., 3â4 in., and 1â2 in. (25 mm, 19 mm, and 12.5 mm). A single 1 in. (25 mm) maximum aggregate size gradation was encountered in the literature which was from Great Britain. This gradation is illustrated in Figure 18. According to the gradation band, the gradation is gapped near the No. 4 sieve and the allowable filler content is between 3.5 and 5.5 percent. Similar to what was found with the survey of agencies, most of the gradation requirements specified by international agen- cies are for a 3â4 in. (19 mm) maximum aggregate size grada- tion. Figure 19 shows the various gradation bands for 3â4 in. maximum aggregate size PFC mixes. Again, customary U.S. sieves are shown on the figure. Also shown on this figure is the gradation band recommended by the National Center for Asphalt Technology (40). This gradation band is shown to provide a comparison between the typical PFC gradation used in the Unite States (as described earlier) and those used in other countries. As shown on this figure, there is a wide range of allowable gradations for PFCs. For instance, on the 3â8 in. (9.5 mm) sieve, gradation requirements range from a high of approximately 75 percent passing (Spain) to a low of approximately 10 percent passing (Italy). However, recall that all of these gradations are included within mix design methods that specify a minimum of 18 percent air voids or more. The majority of gradation bands would force the aggregate blend to be gapped somewhere between the 3â8 in. (9.5 mm) sieve and the No. 4 (4.75 mm) sieve. Filler contents encountered in the various gradation bands also vary significantly. Italy provides a lower limit of 0 percent passing the No. 200 (0.075 mm) sieve while South Africa allows as much as 8 percent passing the No. 200 sieve. Figure 20 illustrates the single 1â2 in. (12.5 mm) maximum aggregate size gradation encountered in the literature review. This gradation band is specified in Great Britain. According to the figure, this 1â2 in. maximum aggregate size gradation would be gapped on either the No. 4 (4.75 mm) or No. 8 (2.36 mm) sieve. The percentage of filler allowed within this gradation band is between 3 and 6 percent. Several authors indicated that the maximum aggregate size selected for PFC will have an affect on permeability. Ruiz et al. (16) indicated that larger maximum aggregate size gradations result in more permeability. Selection of Optimum Binder Content The philosophy of selecting the optimum binder content for PFC mixes is relatively uniform around the world. However, no specific process or procedure that identified an absolute optimum asphalt binder content was identified. Rather, mix 38 0 10 20 30 40 50 60 70 80 90 100 Sieve Size Pe rc en t P as si ng Britain ¾ in.½ in.3â8 in.No. 4No. 8No. 200 Upper Gradation Limits shown as solid lines, Lower Limits shown as dashed lines. 1 ½ in.1 in. Permeable Friction Course Gradations ~ International Agencies Figure 18. 1 in. PFC gradation requirements from international agencies.
39 0 10 20 30 40 50 60 70 80 90 100 Sieve Size Pe rc en t P as si ng Spain Switzerland Italy South Africa NCAT ¾ in.½ in.3â8 in.No. 4No. 8No. 200 Upper Gradation Limits shown as solid lines, Lower Limits shown as dashed lines. Permeable Friction Course Gradations ~ International Agencies Figure 19. 3â4 in. PFC gradation requirements from international agencies. 0 10 20 30 40 50 60 70 80 90 100 Sieve Size Pe rc en t P as si ng Britain ¾ in.½ in.3â8 in.No. 4No. 8No. 200 Upper Gradation Limits shown as solid lines, Lower Limits shown as dashed lines. Permeable Friction Course Gradations ~ International Agencies Figure 20. 1â2 in. PFC gradation requirements from international agencies.
design methods generally identify a range of allowable asphalt binder contents from which the absolute optimum can be selected. Two properties generally are utilized to define the range of allowable binder contents: durability and draindown potential. It should be stated, however, that the mix design methods also require a minimum air void content. Figure 21 illustrates the general concept for selecting the allowable range of asphalt binder contents from which opti- mum is selected. Within this figure, durability is defined as the amount of loss from the Cantabro Abrasion test. This test evaluates the resistance of compacted open-graded specimens to abrasion. The test method entails compacting mix to the laboratory standard compactive effort, allowing the specimen to cool to room temperature, weighing the specimen to the nearest 0.1 g, and then placing the specimen into a Los Angeles Abrasion machine without the charge of steel spheres. The Los Angeles Abrasion machine then is operated for 300 revo- lutions at a rate of 30 to 33 rpm. After the 300 revolutions, the specimen is removed and again weighed to the nearest 0.1 g and the percent mass loss determined based upon the origi- nal specimen mass. This test method was developed in Spain during the 1980s (21). Within the literature, this is the most common test utilized to evaluate the durability of PFCs. As shown in Figure 21, the Cantabro Abrasion test is used to identify a minimum asphalt binder content. As asphalt binder content increases, durability is improved. A maximum asphalt binder content is identified by conducting some type of draindown potential test; more asphalt binder improves durability, but too much asphalt binder leads to draindown. The Cantabro Abrasion test is the most common test uti- lized worldwide to evaluate the durability of PFC mixes. Each country specifying the Cantabro Abrasion test utilizes the same test method with regards to the number of revolutions and rate of revolution within the Los Angeles Abrasion machine. The only variable identified within the Cantabro Abrasion test is the temperature at which the test is conducted. Spain and Belgium utilize a test temperature of 64°F (18°C) (21, 35) and a test temperature of 68°F (20°C) is used in France (44). The remaining countries specify a test temperature of 77°F (25°C). Criteria for the Cantabro Abrasion test are specified based upon the type of conditioning to which the samples are sub- jected. There are three different conditions in which samples are tested: unaged, aged, and moisture conditioned. Specifica- tion values for the Cantabro Abrasion test conducted on un- aged samples are predominantly a maximum percent loss of 25 percent. However, the Texas DOT specifies a maximum of 20 percent loss as does Belgium (35). All other agencies specify a maximum of 25 percent loss on unaged specimens. To reach the aged condition, samples are placed within a forced draft oven at a given temperature for a specified amount of time. Mallick et al. (36) recommended aging samples at 140°F (60°C) for 7 days prior to testing. After aging, the sample was allowed to cool to the Cantabro Abrasion test temperature of 77°F (25°C). Criteria for samples aged in this manner are a maximum of 30 percent loss. The final conditioned state is moisture conditioning. This is practiced in South Africa (7), Italy (7), Great Britain (26), and Australia (19). To moisture condition samples, specimens are submerged in water for a specified amount of time. The only conditions provided in the literature were from Great Britain where specimens are submerged for 24 hours in a 140°F (60°C) water bath (26). It should be stated that all of the references above utilized test samples that were compacted with a Marshall hammer. Watson et al. (40) developed recommendations for Cantabro Abrasion loss values for samples compacted in a Superpave gyratory compactor. In an unaged condition, abrasion loss should be less than 15 percent. There are a number of methods for evaluating the drain- down potential of PFC mixes. Decoene (18) described two methods utilized in Belgium: a basket drainage test and the Schellenberger drainage test. During the basket drainage test, PFC mixes are first compacted in Duriez molds under a pres- sure of 435 psi (30 bars). The molds containing the com- pacted PFC then are placed into an oven maintained at 356°F (180°C). Samples are held at this temperature for 7.5 hours. At the conclusion of this test, the percent asphalt binder lost from the samples is calculated as a percent of the initial binder content (18). The Schellenberger drainage test begins by placing 1,000 to 1,100 grams of loose PFC into a glass beaker. The beaker then is placed into an oven maintained at 338°F (170°C) for 1 hour. After the allotted time, the loose PFC is removed from the beaker and the amount of asphalt binder remaining in the beaker is determined. Draindown potential is described as the binder remaining in the beaker and is expressed as a percentage of the initial asphalt binder content (18). Santha (10) of the Georgia DOT described the Pyrex bowl method for evaluating draindown potential. For this method, 40 Cantabro Loss, % (Du rability) Asphalt Binder Content Dr ai n do wn P ot en tia l Cantabro Abrasion Loss Draindown Potential Maximum Binder Content Minimum Binder Content Cantabro Loss, % (Du rability) Dr ai n do wn P ot en tia l Figure 21. Philosophy of designing PFC mixes.
mix is prepared and placed into a clear Pyrex bowl. The bowl then is placed in an oven set at 250°F (121°C) for 1 hour. A visual examination of the bowl is conducted after the 1 hour to qualify the amount of asphalt binder left in the Pyrex bowl. Santha (10) also states that the Schellenberg drainage has been utilized by Georgia DOT. In a subsequent paper to Santhaâs (10), Watson et al. (11) indicated that the Georgia DOT had adopted the draindown test developed at the National Center for Asphalt Technology. Mallick et al. (36) describe this method as placement of loose mix into a wire basket. The mix and basket are placed into an oven set at the specified temperature. Mallick et al. (36) used test temperatures of 320 and 338°F (160 and 170°C, respec- tively) though later recommendations from the same authors were to conduct testing 27°F (15°C) higher than anticipated production temperatures (4, 39). Within the oven and under- neath the wire basket, a suitable container of known mass is placed. The mix then was held at the elevated temperature for 1 hour. At the end of 1 hour, the basket is removed from the oven and the mass of the container is determined. Draindown then is calculated based on the mass of binder rather than drains from the mix through the basket into the container, and expressed as a percentage of the total mix mass. In a later research project, Watson et al. (40) conducted draindown tests of various PFC mixes using the draindown basket, but with different size wire mesh to fabricate the baskets. The two mesh sizes represented a No. 4 (4.75 mm) screen mesh and a No. 8 (2.36 mm) screen mesh. The stan- dard mesh size was the No. 4 screen. The smaller mesh size was investigated because Watson et al. (40) believed that some intermediate-sized aggregates could pass through the No. 4 sized screen. Another modification to the standard procedure described was that asphalt binder remaining on the basket after the 1 hour was considered as part of draindown. Results of comparisons between the standard draindown test and the modified versions showed very strong correlations. However, Watson et al. (40) stated that tests conducted with the No. 8 (2.36 mm) mesh-sized basket resulted in more repeatable test results. They did not recommend changes to how draindown was determined. During the survey described in Chapter 2, the majority of respondents indicated that draindown testing is included within their mix design methods. Approximately 65 percent of the agencies stated that they utilized the draindown basket method. The remaining agencies utilize the Pyrex bowl method or other technique (not specified). As defined in this report, PFCs are designed to have air void contents greater than 18 percent. As such, a standard labora- tory design compactive effort is needed during mix design. The literature presented two laboratory compaction methods prevalent in designing PFC: the Marshall hammer and Super- pave gyratory compactor. Historically, the Marshall hammer has been used to design PFC mixes. The Marshall hammer has been utilized in Belgium (18), Georgia (10), United Kingdom (7), Spain (7), Italy (7), South Africa (7), and Switzerland (35). Not all references reported the compactive effort when using the Marshall hammer; however, all that did report the design compactive effort reported 50 blows per face, except one. Santha (10) indicated that 25 blows per face were utilized by the Georgia DOT during design (in 1997). Most of the U.S. agencies that place PFCs are currently uti- lizing a Superpave gyratory compactor. The most common design compactive effort with the Superpave gyratory com- pactor is 50 gyrations; however, McDaniel and Thornton (13), utilized 20 gyrations in Indiana. The 50 gyrations was selected during research that compared densities achieved by 50 blows per face of the Marshall hammer and various design gyration levels (40). Subsequent work by Watson et al. (40) conducted a more comprehensive evaluation to determine the appropri- ate design compactive effort. Within this research, the effect of aggregate breakdown was also evaluated. Watson et al. (40) concluded that the design compactive effort of 50 gyrations was appropriate. Though having different operational characteristics than the Superpave gyratory compactor, Alderson (19) reported that Australia also uses a gyratory compactor to design PFCs. In Australia, 80 gyrations of the Australia gyratory compactor are used to design PFCs. Mallick et al. (36) utilized a laboratory permeability test during mix design. The permeability device was described as a falling-head permeameter that was based on an apparatus developed by the Florida DOT. Mallick et al. (36) stated that the laboratory test was optional during the mix design, but indicated that a minimum value of 330 ft/day (100 m/day) should be utilized. Faghriand and Sadd (45) also utilized per- meability testing to evaluate PFC mixes. A final test recommended during the design of PFC mixes is the dry-rodded test to evaluate the existence of stone-on-stone contact. The concept is similar to that used in the design of SMA and is called voids in coarse aggregate (VCA). Kandhal (4) and Watson et al. (42) have recommended the VCA con- cept in designing PFCs. The method entails first measuring the VCA of the coarse aggregate only using AASHTO T19, Unit Weight and Voids in Aggregates. There is a difference between the two references on the definition of coarse aggregate. Kandhal (4) defines the coarse aggregates as those aggregates coarser than the No. 4 (4.75 mm) sieve while Watson et al. (42) utilize the break point sieve as differing between fine and coarse aggregate. Watson et al. (42) defined the break point sieve as the finest sieve to retain 10 percent or more of the ag- gregate blend. The next step in evaluating stone-on-stone contact is to calculate the VCA of compacted samples. If the VCA of the compacted PFC is less than the VCA of the dry- rodded aggregates, then stone-on-stone contact is achieved 41
(4, 42). Watson et al. (42) further verified the existence of stone- on-stone contact using X-ray Computed Tomography. Performance Testing The predominant type of performance testing conducted during PFC designs is moisture sensitivity testing. As men- tioned under the Cantabro Abrasion Loss discussion above, moisture conditioning of PFC samples prior to testing has been utilized (7, 19, 24). To moisture condition samples prior to Cantabro testing, samples are submerged in a heated water bath for a specified amount of time. The most predominant method found in the literature for conducting moisture susceptibility testing on PFCs is to use indirect tensile strength testing and tensile strength ratios (TSR). This also was found to be true in the survey described in Chapter 2. The conditioning of samples prior to determin- ing TSRs varies within practice. Some have recommended the use of five freeze-thaw cycles prior to testing (4, 46), while some agencies responding to the survey indicated that one freeze- thaw cycle was included during TSR testing. In 2004, Watson et al. (39) compared TSR results after 1, 3 and 5 freeze-thaw tests. Results from comparisons showed no significant differ- ence in TSRs after 1, 3, and 5 freeze-thaw cycles. Within the survey of states, the next most common moisture susceptibility test was the boil test. Santha (10) also utilized this test method. This test method essentially entails placing loose mix into boiling water for a specified time. After boiling, a qualitative evaluation of the amount of binder that has stripped from the aggregates is made. The final test identified for evaluating moisture suscep- tibility was a loaded-wheel tester. Cooley et al. (43) loaded samples submerged under water to evaluate moisture suscep- tibility. The loaded-wheel tester used was an Asphalt Pavement Analyzer. A number of other tests were identified in the literature to evaluate designed PFC mixes. In the Netherlands, a dynamic bending test was used to evaluate the stiffness of PFC mixes (17). No specifics were provided on the test, but it is assumed to be similar to the four-point bending beam fatigue test. Addi- tionally, the Netherlands have used a wheel-tracking device to evaluate the rutting performance of PFC mixes (17). Simi- larly, Mallick et al. (36) used the Asphalt Pavement Ana- lyzer to evaluate the stability of PFC mixes. Spillemaeker and Bauer (46) discussed a rotary shearing press to evaluate rut- ting potential. No specifics were provided for this test other than providing the French Standard (NF P 98-252). Another method of evaluating the potential for rutting potential was described by Fortes and Merighi (44). These authors described results from a static, unconfined creep test. Again, specific test conditions were not given. The final two performance tests identified in the litera- ture were reported by Molenaar and Molenaar (38). Both of these tests were designed to evaluate the potential for short- term raveling. The first test was called the Wheel Fretting Test (WFT). For the WFT, a treaded tire inflated to 87 psi (600 kPa) and loaded to 675 lb (3kN) was run in a circular path on top of PFC test specimens. The loaded tire had an inclination angle of 2 to 5 degrees. A total of 3 million revolutions were applied to the test samples at a test temperature of approximately 68°F (20°C). The fretting performance was characterized as a mass loss after the wheel passes. The second short-term raveling test described by Molenaar and Molenaar (38) was called the California Abrasion Test, which utilizes a mechanical shaker that is operated at 20 cycles per second with a specified vibration amplitude. A sample of PFC was placed into a container along with water and steel spheres and subjected to the vibration action for 15 minutes at a test temperature of 39°F (4°C). Again, test results are reported as a percent mass loss after the 15 minutes of abrasive action. 42
