Annotated Literature Review for NCHRP Report 640 (2009)

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.

74 1.15.6 Rehabilitation Practices No information on rehabilitation of porous asphalt mixtures has been provided. 1.15.7 Performance No information on performance of porous asphalt mixtures has been provided 1.15.8 Structural Design No information on structural design of porous asphalt mixtures has been provided. 1.15.9 Limitations No information on limitations of porous asphalt mixtures has been provided. 1.16 Choubane, B., J. A. Musselman, G. C. Page. “Forensic Investigation of Bleeding in Open-Graded Asphalt-Rubber Surface Mixes.” TRB 1999 Annual Meeting CD-ROM, Transportation Research Board. National Research Council. Washington, D.C. 1999. 1.16.1 General In this paper, Choubane et al provides a description of asphalt binder drainage or “bleeding” problems noted in open-graded friction courses (OGFCs) in Florida, possible reasons for the problems and a description of a detailed investigation of such a problem. The authors mention that the use of ground tire rubber has enabled Florida DOT (FDOT) to use more asphalt binder resulting in thicker asphalt binder films, and hence, improve their durability. However, they mention that “fat spots’ resulting from asphalt binder bleeding has still remained a persistent problem. Through a detailed investigation of production and laydown of an OGFC mix, the authors contend that part of the bleeding problem can be traced back to inconsistencies in binder contents in the mix during production, especially during start up in the plants, flushing of surface being overlaid with OGFC and build-up of tracked tacked material. The authors mention that the flushing and build-up of tacked materials go unnoticed during night paving. Choubane et al concludes that fluctuations in binder content during start up could have caused bleeding problems and that lowering the binder content slightly below the optimum design content could help in avoiding significant problems with the mix. They contend that flushing of surface and build-up of tacked material could also lead to the bleeding problem. They recommend that standard start up procedures, as developed in this study, should be followed by the contractor and paving sites should be reviewed in day-light prior to night paving, in order to detect and remove any areas of excessive or tracked tack material. They also recommend that, to avoid design problems associated with OGFCs, dense-graded friction courses be used (in lieu of open-graded friction courses) wherever possible.

75 1.16.2 Benefits of Permeable Asphalt Mixtures In their discussion on the rationale for using OGFC mixes in Florida, Choubane et al mentions that the benefits of using OGFC include reduced hydroplaning potential through greater macrotexture and enhanced drainage of water from the tire/pavement interface resulting in improved tire contact with the pavement, and increased visibility through reduced splash and spray from tires. 1.16.3 Materials and Design Choubane et al provides a description of the FC2 type OGFC mixes used in Florida. Table 42 summarizes the materials and mix design criteria. Table 42: Materials and Mix Design Criteria for OGFC Mixes Used in Florida Material/Mix Type/Property Aggregate Crushed granite, slag, or oolitic limestone as the coarse aggregate component. A typical blend of materials for an FC-2 would include approximately 92% of a No. 89 Stone, and 8% sand or screenings. Gradation Sieve Size Percent Passing 12.5 mm 100 9.5 mm 85-100 4.75 mm 10-40 2.0 mm 4-12 75 µm 2-5 Asphalt Binder AC-30 blended (wet process) with 12% GTR (by weight of asphalt cement Binder Content 7.1% for mixes containing oolitic limestone, and 6.3% for granite and slag mixes 1.16.4 Construction Practices In describing typical OGFC mixes used in Florida, Choubane et al mentions that these mixes are spread at a rate of 27 - 34 kg/m2 , and then “seal rolled” with a tandem steel- wheel roller with a weight not exceeding 2.4 kg/mm of drum width. They also mention that there is no density requirement for OGFC mixes and the acceptance of open-graded friction courses at the plant is based on binder content and extracted gradation. The binder content acceptance is based on a meter/printer system, since the results of binder extraction could be affected by the presence of ground tire rubber (GTR) present in the mix. 1.16.5 Maintenance Practices No information on maintenance practices of porous asphalt mixtures has been provided. 1.16.6 Rehabilitation Practices No information on rehabilitation of porous asphalt mixtures has been provided.

76 1.16.7 Performance Choubane et al mentions several cases where bleeding has been observed, as well as a case where a thorough investigation was made to determine possible causes of bleeding. These projects and their performances/distresses, in terms of durability/bleeding are summarized in Table 43. Table 43: Projects with Bleeding Problems Project Details Performance/Distress SR-64 Manatee County Placed during the summer of 1995; The aggregate blend consisted of 88% oolitic limestone and 12% limestone screenings. The total binder content was 7.1%. Following construction, flushing was observed at three distinct pavement locations, in less than 5% of the total length of the project. During production, the meter/printer system indicated the binder content was close to the target and gradations of the extracted aggregate also matched the mix design targets. A subsequent investigation indicated that the bleeding in the distressed areas resulted from a low binder viscosity caused by settlement of the GTR in the asphalt storage tank, which resulted in draindown during construction. I-75, Pasco County Placed during the summer of 1995. The aggregate blend consisted of 92% Nova Scotia granite and 8% sand, with a total binder content of 6.5%. The project experienced flushing in several locations throughout its length. The cause of the bleeding was never clearly established. During production, the meter/printer system indicated the binder content was close to the target and gradations of the extracted aggregate also matched the mix design targets. The mix design was re-verified and the results indicated that the original binder content was appropriate. Furthermore, a similar mix design was placed on an I-75 project located in nearby Sumter County which did not experience any bleeding. SR-54, Pasco County and SR-600, Hillsborough County Placed during the spring of 1996. Aggregate blend consisted of 94% oolitic limestone and 6% sand, with a total binder content of 7.1%. Both projects developed numerous areas of bleeding. A re- verification of the mix design showed that the optimum binder content of the mix had dropped from 7.1% when it was originally designed in 1994, down to 6.4%. The reduction in the binder content was related to a change in aggregate surface texture, apparently caused by a change in the crushing operation at the quarry. A review showed that out of 24 open-graded mixes utilizing aggregate from this source, the only two which experienced bleeding problems were the SR-54 and SR-600 projects. In view of these projects (all in the Tampa area), Choubane et al mentions that FDOT decided to re-verify all OGFC mix designs to be used on projects in the Tampa area prior to their use, and issue these mix designs at a binder content 0.2% lower than optimum. These were done to prevent the plant produced OGFC from changing significantly from its mix design and hence to prevent resultant durability problem. However, the authors mention that in the summer of 1997, another project in the Tampa area developed intermittent bleeding project, which prompted an investigation. The details of this project, the investigation and follow-up work are summarized in Table 44.

77 Table 44: Investigation and Follow-Up Work Mix Design/Construction Distress/Cause Follow-up work Aggregate blend: 92% Nova Scotia granite and 8% sand. The optimum binder content of the mix was 6.0%. The mix was placed on this project during the early summer of 1997. Within four to six weeks after construction, several areas of the project appeared to have excessive asphalt that had flushed to the surface. There were approximately 5 locations within the project that had bleeding ranging in length from 3 to 30 m (10 to 100 feet). In most instances, the bleeding occurred within the wheel-paths; the bleeding seemed to be more severe at a major intersection within the project. The intermittent occurrence of the bleeding throughout the project seemed to indicate that the problem was not exclusively related to the mix design (confirmed with reverification of mix design). 1. FDOT redesigned the open- graded friction course mix, substituting limestone screenings for the sand. It was believed that the screenings would increase the overall absorption of the mix, thus minimizing the potential for draindown. The total binder content was 5.9% (which was approximately 0.5% below the optimum determined by the open-graded friction course design procedure.) 2. The production of the open- graded friction course mix was monitored at the contractor’s plant. The asphalt content for each sample was determined using the ignition method 3. Core samples were tested for binder content, gradation, and thickness. Choubane et al indicates the following observations from the follow-up work: 1. Binder content from first night’s production showed excessive fluctuations in binder contents (7.09, 5.94, & 6.82 with a target of 5.90). The binder content data obtained for the second night was more consistent, with only one relatively high test result. 2. Test results on cores indicated constant binder contents at most areas but significantly higher binder contents at areas showing bleeding. 3. The course was placed very thin, 6-10 mm, and hence would have the potential of exhibiting more bleeding than usual, because of draindown or build-up of tacked materials. Subsequently, Choubane et al mentions, FDOT monitored another OGFC paving project, and determined that materials and mix design parameters closely met target values. However, they did note that there was flushing on the surface which was being overlaid and also build-up of tacked material. Choubane et al indicates that concurrent to FDOT’s monitoring operations, the contractor implemented an enhanced method of production of OGFC mixes. This method consists of: allowing the drum mixer to run empty (clean-out) for 5 minutes; tighter documentation on material usage; closer visual observation of the mix during production; constant/maximum production rates for each plant; and testing the initial mix as it is produced. In addition, Choubane et al indicates that sampling and monitoring (and testing of GTR and AC 30) were also conducted during the GTR blending operation in the asphalt rubber terminal in Tampa. Testing of materials was conducted at different stages – start from the