Validation of a Performance-Based Mix Design Method for Porous Friction Courses (2021)

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.

3 C H A P T E R 2 State of Practice Findings of NCHRP Project 01-55 Survey In 2015, NCAT conducted a survey of state agencies to determine those that were using PFC mixtures, the mix design procedures used, the primary forms of distress for those mixes, and any improvements that had been made to address the observed distresses. Results from that survey were included in NCHRP Research Report 877, Performance-Based Mix Design of Porous Friction Courses, (Watson et al., 2018). Key findings from the survey are summarized below. To improve PFC performance, revisions to mix design test procedures and construction practices were made. Reported Distresses While PFC mixtures have several inherent safety and environmental benefits, half of the responding agencies do not use PFC mixtures because of performance issues. Seventy-five percent of the respondents reported that raveling was the leading form of distress that resulted in reduced service life. Other performance issues included cracking, moisture damage, and loss of functionality (clogging). The performance life of the PFC mixture generally ranged from 6-10 years with only a few states indicating a typical life beyond 10 years. In addition, 43 percent of the respondents indicated they did not have performance test requirements addressing the observed distresses. The Cantabro procedure (AASHTO TP 108) is used by 24 percent of the respondents for assessing raveling potential; 33 percent use a modified version of AASHTO T 283 for evaluating resistance to moisture damage. PFC Mix Design An asphalt mix design for any mix type generally consists of at least four steps: (1) select quality material, (2) determine suitable blend gradation, (3) determine optimum asphalt content, and (4) evaluate potential performance. Both AASHTO and ASTM have published PFC mix design procedures (viz., AASHTO PP 77 and ASTM D7064), but an observation from the NCAT survey was that all agencies use an agency- specific design procedure or some other method that may only include selected parts of the AASHTO and ASTM methods. For example, several agencies use a surface capacity test to determine optimum asphalt content, some use a Pyrex pie plate method, and some select optimum based on laboratory air voids. The Pyrex pie plate results may be problematic in that they depend on asphalt binder draindown (and most agencies currently use polymer modified binder and/or fiber stabilizer to prevent binder draindown), which can result in a very high optimum binder content. The main purpose of PFC is to drain rainfall through the mixture void structure rather than across the surface. This reduces the potential for hydroplaning, reduces backspray, and reduces headlight glare from approaching vehicles. However, 95 percent of the respondents indicated they do not have a permeability requirement to evaluate drainage potential during the mix design process. Several agencies have revised their mix design procedures to incorporate some of the test requirements recommended in the AASHTO and ASTM methods, as well as recommendations from national- and state-

4 sponsored research. The changes are generally related to materials selection, laboratory performance test procedures, and construction practices, as discussed below. Materials Selection Aggregate. The aggregate component of the mixture primarily consists of coarse aggregate in order to provide stone-on-stone contact and an inter-connected void structure. Granite and limestone are the predominate aggregate types used in PFC. Most agencies have specifications such as polished stone value, Los Angeles abrasion, flat and elongated particles, and soundness for coarse aggregate quality. The proportion of fine aggregate is limited in order to avoid filling the void structure and reducing the drainage capacity of the mix, but a small proportion is needed to help provide cohesion. Fine aggregate quality is typically evaluated by tests for sand equivalent, soundness, and uncompacted voids. Asphalt Binder. The asphalt binder is typically performance graded using the Superpave binder testing procedures although some foreign countries still use a penetration grading system. Almost all agencies require the binder to be modified with either polymer or asphalt rubber to meet the PG standards although some success has been achieved without modification. These modifiers can provide a stiffer asphalt binder for PFC mixtures, which leads to an increase in cohesion of aggregate stone skeleton. For this reason, the PFC mixes with modified asphalt binders usually have higher resistances to rutting, cracking, and raveling damage, which exhibit better durability in the field. Design Gradation. A 12.5 or 9.5 mm nominal maximum aggregate size mix is often specified with the gradation ranges closely following earlier FHWA recommendations or NAPA IS-115 guidelines. The 12.5- mm PFC maximizes the water draining capacity of the mix and is especially useful if water is drained across several lanes. The 9.5-mm mix is more useful for noise reduction and is more economical because it can be placed in thinner layers. The 12.5-mm NMAS mix has been more prevalent in the U.S., but some states such as GDOT and SCDOT are beginning to place more 9.5-mm mix. For example, SCDOT revised specifications in 2019 to include a finer 9.5-mm PFC as well as their 12.5-mm PFC. A table summarizing the gradation limits specified by state agencies is included in NCHRP Research Report 877 (Watson et al., 2018). Laboratory Test Procedures Cantabro Stone Loss. Very few states were using a performance test during the mix design phase that would indicate susceptibility to the types of distress encountered on constructed projects. Several PFC mix design procedures, including NAPA IS-115, AASHTO PP77, and ASTM D7064, have recommended the Catabro test procedure for evaluating mixture durability. Although the test is empirical in nature, it is widely recognized in PFC research. SCDOT and TxDOT (Tex 245-F) have already implemented the Cantabro stone loss procedure (AASHTO TP 108) with a maximum loss of 15 percent for South Carolina and 20 percent for Texas. Recent research has recommended a maximum loss of 15 percent for Alabama (Watson et al., 2020).

5 Moisture Test. The large volume of water channeled through a PFC layer increases the potential for stripping of the asphalt from the aggregate particle. As a result, most agencies require hydrated lime or other anti-stripping additives to prevent moisture damage. A modified version of AASHTO T 283 is typically used to evaluate moisture susceptibility. Modifications include compacting loose mix to 50 gyrations with a Superpave gyratory compactor, vacuum saturating compacted specimens for 10 minutes, and keeping compacted specimens submerged during the freeze-thaw cycle. The NCHRP Project 01-55 performance-based mix design procedure proposed a minimum conditioned tensile strength of 50 psi and a minimum tensile strength ratio (TSR) of 0.70. Rather than using a modified approach to AASHTO T 283, TxDOT uses the HWT Test (Tex 242-F). Fine graded PFC mixtures must meet a minimum of 10,000 passes before exceeding 12.5-mm rut depth. The test is performed for informational purposes only for their coarse graded PFC. Cracking. TxDOT is the only state that requires a performance test to evaluate susceptibility to cracking. The Texas Overlay Tester (Tex 248-F) is specified, and the fine graded PFC mix must withstand a minimum of 200 cycles before failure. Construction and Maintenance Issues during construction may be more responsible for poor performance of PFC than mixture properties (Bennert et al, 2014). Several agencies have indicated that increasing the tack rate and using a polymer modified tack coat have improved PFC performance. ALDOT has practically doubled the previously required target tack application rates of 0.05 to 0.10 gal/sy for emulsion tack (and 0.03-0.07 gal/sy for PG tack). ALDOT revisions in 2018 limit tack material to CQS-1hP emulsion of polymer modified binder, and the revised application rates are 0.18-0.23 gal/sy for CQS-1hP emulsion and 0.13-0.18 gal/sy for PG tack. SCDOT now requires only hot applied less-tracking bond coats or PG 64-22 binder for tack coat, and the application rate is a minimum of 0.08 gal/sy. Placing cold mix or placement in cold temperatures may also reduce PFC performance life. Raveled areas beginning at transverse joints and where the paver was stopped during construction have been a problem experienced by several agencies. This may be caused by the first load of mix produced at start-up of the plant not being a consistent temperature throughout. It may also be due to cooling if the metal components of the plant, material transfer vehicle (MTV), or paver are not sufficienlty heated prior to paving. SCDOT has addressed this issue by allowing the contractor to produce half a load of PFC at the beginning of each lot to be run through the MTV and paver in order to sufficiently heat the roadway equipment at start-up. That material is discarded. Another approach is to hold the first load and make the transverse joint with the second or third load. SCDOT also requires the mix be at least 225°F by the time it is finally placed on the roadway. Implementation of Performance-Based Mix Design A product of NCHRP Project 01-55 was a proposed AASHTO mix design procedure included in the Appendix of NCHRP Research Report 877 (Watson et al., 2018). The procedure recommended criteria for materials selection and typical PFC gradation bands for 9.5-mm, 12.5-mm, and 19-mm mixtures. Aggregates are required to meet the criteria for Superpave mixtures according to AASHTO M 323. It is recommended that asphalt binder meet requirements for AASHTO M 320 and that the high temperature grade be two grades higher than the standard agency paving grade to account for high traffic volume projects or where slow to standing traffic is encountered. Trial samples are prepared at various asphalt contents and compacted to 50 gyrations with a Superpave gyratory compactor. The determination of optimum asphalt content is based on the following criteria:

6  Design air voids. Air voids may be determined by either the vacuum sealing method (AASHTO T 331) or by the dimensional method. The recommended air void range is 15-20 percent if the vacuum seal method is used, or 17-22 percent if the dimensional method is used.  Maximum Cantabro stone loss. The Cantabro stone loss is determined using AASHTO TP 108. Unaged specimens are typically used. Figure 2-1 shows the effect of design air voids (AASHTO T 331) on the Cantabro loss for the mixtures tested in NCHRP Project 01-55 (Watson et al., 2018). The proposed maximum Cantabro loss for the recommended air void range of 15-20 percent was originally set at 15 percent, but it was suggested later to change to 20 percent so that the same threshold could be used for both mix design and production acceptance. Some agencies currently require a maximum threshold of 15 percent stone loss based on AASHTO PP 77-14 or results from earlier research studies.  Minimum shear strength. A Marshall Stability test apparatus can be used with a test fixture developed to determine bond strength between pavement layers to apply the shear load for this test. The shear stress is an indication of the cohesiveness of the material and is calculated by dividing the peak load by the cross-sectional area of the specimen. A minimum shear strength of 125 psi is recommended.  Minimum permeability. The Florida permeability method FM 5-565 is used to determine permeability, and a minimum threshold of 50 m/day is recommended. (a) (b) (c) (d) Figure 2-1. Effect of Specimen Air Voids on Cantabro Loss (Watson et al., 2018).

7 In addition, I-FIT can be used according to Illinois Test Procedure 405 for optional cracking evaluation with a proposed minimum FI value of 25, and HWT can be used for the optional rutting susceptibility test with the following proposed number of cycles to failure (12.5 mm rut depth). – PG 64 Min. 10,000 passes – PG 70 Min. 15,000 passes – PG 76 or higher Min. 20,000 passes Case Studies of Implementation Several agencies are reviewing their specifications and mix design procedures as a result of recent research results regarding PFC mixtures. Changes made, or being considered, by South Carolina, Georgia, Tennessee, and Alabama are used as case studies in this report. South Carolina SCDOT has been the most active in making revisions to its specifications for PFC mix design, production, and construction improvements. The latest changes were published in a January 2019 supplemental specification. Materials Selection. SCDOT prohibits the use of RAP, slag, or marine limestone in PFC mixtures. Hydrated lime is required as an anti-stripping agent and polymer modified PG 76-22 binder is required. A terminal-blended WMA chemical additive is required and eliminates the need for fiber stabilizing additives. Mix Design. A 9.5-mm PFC has been added to the 12.5-mm PFC typically used in the past. The gradation ranges are shown below: Table 2-1. SCDOT PFC Gradation Limits. Sieve Size Percent Passing Std Metric 12.5 mm PFC 9.5 mm PFC 3/4 inch 19.0 mm 100 100 1/2 inch 12.5 mm 85 – 100 95 – 100 3/8 inch 9.5 mm 55 – 75 80 - 100 No. 4 4.75 mm 15 – 30 20 – 50 No.8 2.36 mm 5 – 15 5 – 20 No. 200 0.075 mm 0.0 – 4.0 0.0 – 4.0 The Cantabro abrasion loss is required to be less than 15 percent. Asphalt binder content must be within a range of 5.5 – 7.0 percent with a minimum retention coating of 99.5 percent. Rather than air voids, SCDOT specifies a minimum of 13 percent porosity using the test procedure SC- T-128, Porosity of Compacted Open Graded Friction Course (OGFC) Mixture Specimens. Production. To avoid cross-contamination, the plant is not permitted to produce other mixes while producing PFC mixture. Production temperature must be at least 225°F and mix must be placed within 90 minutes of being discharged from the plant.

8 Construction. Only a less-tracking hot applied tack coat, or PG 64-22 binder, may be used for tack coat. The minimum tack application rate is 0.08 gallons/square yard. To reduce the potential for raveling at transverse joints, SCDOT allows one-half load of PFC to be sent to the project to preheat the MTV and paver at start-up before being discarded. When mix is placed on the roadway, it must not be less than 225°F. Action must be taken by the contractor to eliminate paver stops. Georgia Georgia has been one of the leading users of PFC mixtures for surface courses on interstate and high traffic volume state routes for the last 30 years. The typical service life is 10-12 years with some mixes lasting up to 18 years. The primary form of distress has been raveling. In the mid 1990s, GDOT implemented use of a Porous European Mix (PEM, based on research reports from Europe) that is coarser than the 12.5-mm PFC previously used. The coarser mix was placed in a 1.25-inch layer and was helpful in providing additional water drainage capacity and was most beneficial for draining water across multiple lanes. However, the PEM mix is more costly to place because of the increased layer thickness (typical PFC mixes are placed 0.75 to 1.0 inches in thickness). GDOT is in the process of revising PFC specifications due to changes in mix design and construction. Mix Design. GDOT is revising PFC (OGFC) specifications to allow more filler (passing the No. 200 sieve) based on NCHRP 1-55 research that showed the additional filler may provide more cohesion within the mix and resist the potential for raveling. The proposed gradation for mixes is shown below. Table 2-2. GDOT PFC Gradation Limits. Sieve Size Design Gradation Limits, Percent Passing Std Metric 9.5-mm PFC 12.5-mm PFC 12.5-mm PEM 3/4 inch 19 mm 100 100 1/2 inch 12.5 mm 100 85-100 80-100 3/8 inch 9.5 mm 85-100 55-75 35-60 No. 4 4.75 mm 20-40 15-25 10-25 No. 8 2.36 mm 8-15 5-15 5-10 No. 200 0.075 mm 4-6 3-5 2-4 Range for AC 6.25-7.25 6.0-7.25 5.75-7.0 Design Air Voids, % 15-18 15-20 18-22 GDOT is implementing a mix design air void range for the first time as shown in the above table. Maximum allowed draindown is 0.3 percent and the minimum TSR is 80 percent. GDOT also plans to require the Cantabro test for information only to collect sufficient data to establish acceptable criteria based on Georgia materials. Construction. Several agencies have reported increasing tack rate in order to improve PFC performance. GDOT allows three tack materials: (a) PG 58-22, 64-22, or 67-22 applied at a target rate of 0.06-0.08 gal/sy, (b) non-tracking hot applied polymer modified tack applied at 0.12-0.18 gal/sy, or (c) CQS special modified emulsion used in conjunction with a spray paver at 0.22-0.28 gal/sy. The target placement spread rate is 75-95 lb/sy for 9.5-mm PFC, 85-110 lb/sy for 12.5-mm PFC, and 110-165 lb/sy for PEM mixtures. In order to reduce the potential for raveling due to cold mix, GDOT requires a continuous paving operation. In the event the paver stops, a transverse joint is required if the mix immediately behind the paver cools to less than 250°F.

9 Tennessee The Tennessee Department of Transportation began reuse of PFC mixtures about four years ago, and generally follows NAPA IS-115 guidelines for mix design. Tennessee is somewhat unique in that TNDOT allows either the 50-blow Marshall method or 50-gyration method for laboratory compaction of mixtures. The Marshall procedure is widely used throughout the rest of the world, but the Superpave gyratory compactor is typically used in the U.S. TNDOT also specifies one of the highest air void requirements with design air voids of at least 20 percent. The Cantabro test for mixture durability of unaged specimens must have no more than 20 percent stone loss. A modified version of ASTM D4867 is used for determining resistance to moisture damage. The minimum TSR is 80 percent and the minimum tensile strength is 50 psi. Alabama ALDOT has a long history of using PFC mixtures for interstate and high traffic volume routes, but due to problems with mixture raveling, ALDOT currently limits the use of PFC to selected projects. Still, ALDOT has taken several steps intended to improve the service life of these mixes. The changes implemented have been related to mixture components and construction practices. Mixture Components. The 2012 ALDOT specifications allowed up to 10 percent RAP in PFC mixtures, so the RAP binder generally replaced about 0.6 percent of the asphalt binder. ALDOT requires PG 76-22 asphalt binder grade for PFC mixes so the net effect of adding RAP is that about 10 percent of the binder was not polymer modified. ALDOT revised the specifications in 2018 to eliminate the addition of RAP in PFC mixtures. The range for asphalt content in PFC has been 4.7 to 9.0, but in 2020 a special provision was implemented that raised the minimum asphalt content to 6.0. ALDOT also requires the use of either fiber stabilizers or WMA technology (Evotherm 3G) to prevent binder draindown. Construction Practices. The effort to minimize raveling has also been extended to roadway operations primarily by increasing tack coat quantities. In 2012, ALDOT specifications required 0.05-0.1 gal/sy for emulsion tack and 0.03-0.07 gal/sy for PG tack. Specification revisions in 2018 significantly increased the tack coat spread rate to 0.18-0.23 gal/sy for CQS-1HP emulsion or 0.13-0.18 gal/sy for PG tack. If emulsion is used, it must be applied with a spray paver.