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.

52 1.11.8 Structural Design No information has been provided on structural design. 1.11.9 Limitations No information has been provided on limitations. 1.12 Santha, L. “A Comparison of Modified Open-Graded Friction Courses to Standard Open-Graded Friction Course.” FHWA-GA-97-9110. Georgia Department of Transportation. Forest Park, Georgia. April 1997. 1.12.1 General This report documents a research study conducted by the Georgia Department of Transportation (GDOT) to determine the improvements made to existing GDOT OGFC mixes. Test sections of seven OGFC mixtures were constructed on Interstate 75 south of Atlanta, Georgia. The average daily traffic (ADT) for the interstate was 47,000 with approximately 21 percent truck traffic. The existing pavement prior to construction of the test sections consisted of 140mm of HMA overlying a 9 in. Portland cement concrete pavement that was jointed at 30 ft intervals. In the test sections, 54mm of existing pavement was milled and 35mm of dense-graded surface HMA was then inlaid. The OGFC test mixtures were then placed at a thickness of 19mm over the dense-graded surface mixture. Each of the seven test sections were constructed approximately 0.8km in length. The following lists the seven mixtures placed during the study: Test Mixture Description d Standard OGFC D Coarse OGFC (after improvements) D16R Coarse OGFC with 16% Crumb Rubber DM Coarse OGFC with mineral fibers DC Coarse OGFC with cellulose fibers DP Coarse OGFC with SB polymer DCP Coarse OGFC with SB polymer & cellulose fibers Based upon the results of this study, Santha concluded the following: 1. “The use of coarse OGFC should be continued and the use of fine OGFC discontinued.” 2. “Both fiber and polymer should be included in coarse OGFC. Neither additive appears to perform significantly better or worse than the other in coarse OGFC.” 3. “The use of crumb rubber in coarse OGFC is not recommended.” 1.12.2 Benefits of Permeable Asphalt Mixtures Benefits mentioned by the author included improved wet weather driving conditions which includes reduced hydroplaning and splash/spray and improved pavement friction and surface reflectivity. 1.12.3 Materials and Mix Design Historically, GDOT had used fog seals to help hold aggregates within OGFC when raveling problems had arisen. Fog seals fill surface voids within OGFC and, thus, reduce permeability.

53 In an effort to better maintain permeability, GDOT began to require a larger nominal maximum aggregate size gradation for OGFCs. The GDOT specifications also required the addition of a thermoplastic modifier and mineral fibers in order to help hold the aggregates in-place and to eliminate draindown during construction. Table 28 presents the gradation requirements of the standard OGFC and coarser OGFC used by GDOT (and this research study). Table 28: GDOT OFGC Gradation Requirements Percent Passing Sieve Size, mm Standard OGFC Coarse OGFC 19.0 --- 100 12.5 100 90-100 9.5 85-100 65-85 4.75 20-40 15-25 2.38 5-10 5-10 0.075 2-4 2-4 The design of the OGFC mixtures entailed three primary steps. The first step in the mix design process was to determine the surface capacity of the aggregate fraction that is retained on the 4.75mm sieve. To determine the surface capacity, the aggregate fraction was completely immersed in S.A.E. No. 10 oil for five minutes. After draining the oil from the aggregates, the percent oil retained was determined. The oil retained on the aggregates was then used to calculate the required asphalt binder content for the mix. The second step of the GDOT mix design process was a modified Marshall design. Mixture was compacted at varying asphalt binder contents using 25 blows per face with the Marshall hammer. For each mixture, the bulk specific gravity was determined using volumetric calculations and the mass of each sample. Using the bulk specific gravity, the volumetric properties of each sample were determined. A plot was then developed between voids in mineral aggregate (VMA) and asphalt binder content. Optimum asphalt binder content was selected as the binder content that produces the lowest VMA. The final step in the process was to select an asphalt binder content based upon the Pyrex Bowl Method. For this method, mixture was prepared and placed into a clear glass (Pyrex) bowl. The GDOT method states to start with a binder content of 5.5 percent and then repeat the process for asphalt binder contents of 6.0, 6.5 and 7.0 percent. The Pyrex bowls were placed in an oven set at 121˚C (250°F) for one hour. Then, a visual examination of the amount of liquid that drained from the aggregate structure and left on the Pyrex bowl was conducted. An asphalt binder content was selected where ample bonding was evident without excessive draindown. Optimum asphalt binder content for the mixture was selected as the average binder content from the three steps described above. In addition, two other tests were conducted to evaluate the designed mixes: Cantabro Abrasion and the Schellenberg Drainage test. Table 29 summarizes the designed mixes.

54 Table 29: Summary of Mix Design Information of Research Mixes Sample Type Coarse OGFC (D) D+16% Rubber (D16R) D+ Mineral Fibers (DM) D+Cellulose Fibers (DC) D+SB Polymer (DP) DC+SB Polymer (DCP) Sieve Size Total Percent Aggregate Passing By Mass ¾ in. 100 100 100 100 100 100 ½ in. 99 99 99 99 99 99 ⅜ in. 75 75 75 75 75 75 No. 4 18 18 18 18 18 18 No. 8 8 8 8 8 8 8 No. 200 2 2 2 2 2 2 Percent Asphalt Binder in Total Mix % AC 6.0 6.6 6.3 6.4 6.2 6.4 Miscellaneous Test Data Cantabro (% Wear) 13.5 8.6 5.7 5.8 8.6 8.2 Drainage (% Loss) 0.37 0.05 0.06 0.06 0.34 0.04 Film Thickness (µm) 34.07 36.90 35.92 36.54 35.30 36.54 Little information was provided for the materials. The base asphalt binder before any modification was an AC-20 Special. Both hydrated lime and a liquid anti-strip agent were added to the mixture in order to minimize the potential for moisture damage. All aggregates were a granite-gneiss that had a Los Angeles Abrasion of 42 percent. 1.12.4 Construction Practices A drum mix plant was used to produce the OGFC mixes used in the research study. Mix exiting the drum was carried to a silo for depositing into the haul trucks. In order to minimize the potential for draindown, a maximum of 50 tons of OGFC was stored in the silo at any given time. The production rate for the OGFC test mixes was 150 tons/hr, while the production of typical dense-graded mixes was 250 tons/hr. Two modifications to the plant were required in order to produce the OGFC: introduction of the fibers and asphalt rubber. In order to introduce the fibers, a fiber blowing machine was used. This machine took baled fibers, fluffed the fiber, crowded the fiber to a constant density using a series of paddles and augers, and then blew the fibers to the drum using an air jet. Fibers were blown into the drum at the same location as the hydrated lime. This location was selected based upon a preliminary investigation designed to determine the best location. The two location points investigated included the point at which hydrated lime and asphalt binder were introduced. Trial mixes were produced in order to evaluate these two locations. Samples of baghouse fines were obtained and evaluated to determine if the fibers were being caught in the high velocity drum exhaust gases. Evaluation of the baghouse fines showed no appreciable amount of fibers within the baghouse fines when either of the introduction locations were utilized. Tensile strength tests were also conducted on mix produced when using the two introduction points. Santha indicated

55 that tensile strengths were “slightly higher” for OGFC mix produced when the fibers were added at the asphalt binder injection point; however, it was decided to introduce the fibers at the point that the hydrated lime was added. This decision was based upon this location allowing the fibers to be dry mixed with the aggregates for a short time before the asphalt binder was introduced. The fiber hopper used to introduce the fiber into the mixing process was calibrated prior to initiation of production. The fiber was blown into a pre-weighed sealed container for one minute. The rubber modified asphalt binder was mixed on-site using a batch type blending unit. The rubber modified binder was gravity fed into a pump which fed the mixing process. Both the conventional and modified OGFC mixes were placed and compacted with no major problems. The following are excerpts from the report on some minor issues during placement and compaction. “Coarse OGFC with Mineral Fibers (DM). During laydown, this mix behaved as if it were too cold, even though temperatures fell within job mix formula limits. During the lab design process, GDOT observed that the mix could probably be placed at a higher temperature without the risk of draindown. The JMF was set at 275°F, and the truck temperatures taken at the plant averaged 278°F.” “Coarse OGFC with Cellulose Fibers (DC). This mix, which contained cellulose fibers, had a dull, flat appearance. Problems were encountered when the mix stuck to the roller drums and was picked up. Keeping the rollers back to allow the mix to cool before rolling seemed to help this problem, but adding soap or releasing agent to the drum watering system is likely to be the best alternative…” “Coarse OGFC with SB Polymer and Cellulose Fibers (DCP). This mix contained SB polymer and cellulose fibers, and it was difficult to lay so that it matched the adjoining, previously placed lane. The thickness depended on paver speed, and any increase or decrease in speed would cause the mat to get thinner and thicker, respectively…” “Coarse OGFC with SB Polymer (DP). This mix contained Styrelf polymer, and it was the most difficult to lay. Paver speed changes greatly affected the thickness of the mix, as with the DCP mix. Cores taken from this section had thickness variations of up to 0.4 in. AC draindown occurred in four loads, indicating that the JMF placement temperature (300°F) was probably set too high… The difficulty in laying this mix and the DCP mix is reflected in the relatively high Maysmeter values of these mixes compared with those of the other mixes…” “Coarse OGFC with 16% Crumb Rubber (D16R). This mix, which contained 16% crumb rubber, produced only minor laying or compaction problems. Some pulling of the mat occurred for approximately the first 75 feet but disappeared shortly afterward. The paver screed may have been cold, since this mix was the first which was placed that day…” 1.12.5 Maintenance Practices No specific maintenance practices were given.

56 1.12.6 Rehabilitation Practices No specific rehabilitation practices were given. 1.12.7 Performance Each of the test sections were monitored for 3.5 years after construction. Properties of the test sections that were monitored included: friction, smoothness, visual distress, rutting and permeability. Friction testing was conducted in accordance with ASTM E274 at the following time increments after construction: one day, two weeks, six months, and 3.5 years. Table 30 presents the results of this testing. These results show that initial friction values (one day) were relatively low compared to the other time increments. However, the friction numbers increased significantly at the two week increment and maintained relatively high friction numbers for the 3.5 years. Table 30: Average Friction Test Results for Test Sections Friction Number Test Section 10/27/92 11/11/92 4/12/93 2/6/96 Std. OGFC (d) 42 53 52 50 Coarse OGFC (D) 41 50 52 51 D + Mineral Fibers (DM) 39 50 53 49 D + Cellulose Fibers (DC) 37 47 53 49 DC + SB Polymer (DCP) 35 46 52 50 D + SB Polymer (DP) 32 47 51 51 D + 16% Crum Rubber (D16R) 37 48 53 51 At 3.5 years after construction, smoothness testing was conducted using a new laser profiler. Table 31 presents the results of smoothness testing conducted on each of the seven test sections. An acceptable roughness in Georgia for OGFC is 750 mm/km. Results shown in Table 31 indicate that after 3.5 years, all of the sections had acceptable smoothness. Table 31: Average Smoothness Values in Test Sections Test Section Smoothness Value (5/1996), HRI, mm/km Std. OGFC (d) 427 Coarse OGFC (D) 547 D + Mineral Fibers (DM) 564 D + Cellulose Fibers (DC) 551 DC + SB Polymer (DCP) 543 D + SB Polymer (DP) 585 D + 16% Crum Rubber (D16R) 642 Following construction, visual distress surveys were conducted annually on the different test sections. After 3.5 years, the primary visual distresses consisted of reflective cracking and some longitudinal cracking. Table 32 presents the results of the distress survey on test sections.

57 Table 32: Results of Visual Distress Survey on Test Sections Test Section Description of Distresses # Cracks in 0.2 Mile Std. OGFC (d) Low severity reflective cracks 7 Coarse OGFC (D) Medium severity reflective cracks and some longitudinal cracks at right edge of right lane 26 D + Mineral Fibers (DM) Low severity reflective cracks 10 D + Cellulose Fibers (DC) Low severity reflective cracks 3 DC + SB Polymer (DCP) Low severity reflective cracks 12 D + SB Polymer (DP) Low severity reflective cracks 4 D + 16% Crum Rubber (D16R) Low-medium severity reflective cracks & some longitudinal cracks at right edge of right lane 20 Rut depth measurements were taken in each of the test sections at several time intervals up to 3.5 years. Rut depths were measured with a string line at 100 ft intervals in both the left and right wheel path. Table 33 presents the average rut depths for each test section in 1993, 1994 and 1995. This table shows that rut depths were relatively small in all of the sections containing the Coarse OGFC gradation with the different additives. The largest rut depths were observed within the Standard OGFC section after 3.5 years of traffic. Table 33: Average Rut Depths in Test Sections (inches) 11/93 9/94 2/96 Test Section LWP RWP LWP RWP LWP RWP Std. OGFC (d) --- --- --- --- 0.25 0.28 Coarse OGFC (D) 0.09 0.06 0.14 0.17 0.16 0.10 D + Mineral Fibers (DM) 0.17 0.14 0.14 0.19 0.07 0.14 D + Cellulose Fibers (DC) 0.16 0.14 0.13 0.17 0.15 0.17 DC + SB Polymer (DCP) 0.11 0.05 0.05 0.14 0.12 0.06 D + SB Polymer (DP) 0.13 0.03 0.03 0.12 0.18 0.08 D + 16% Crum Rubber (D16R) 0.07 0.04 0.04 0.14 0.16 0.17 An in-place permeability test was used to evaluate the drainage characteristics of the OGFC sections at the time of construction. Permeability tests were also conducted after 1 year and 3.5 years. Table 34 shows the average results of permeability testing conducted on the test sections. Permeability tests were conducted within the wheel paths and between the wheel paths at a minimum of three locations along the length of each test section. This table shows that the permeability of all test sections decreased substantially over time. Santha states that the decrease in permeability was due to clogging.