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.

120 1.27.8 Structural Design Bishop and Oliver mention that currently BC Ministry of Transportation and Highways currently considers a structural strength value of 1.25 (in terms of Crushed Granular Equivalency) for OGFC, as compared to 2.0 for conventional asphalt pavement. 1.27.9 Limitations Although not mentioned specifically, the authors caution against the use of OGFC mixes in areas which would not get early or close attention for winter maintenance because of their location or importance. 1.28 Bolzan, P. E., J. C. Nicholls, G. A. Huber. “Searching for Superior Performing Porous Asphalt Wearing Courses.” TRB 2001 Annual Meeting CD-ROM. Transportation Research Board. National Research Council. Washington, D.C. 2001. 1.28.1 General This paper provides descriptions of design of porous asphalt mixes in the United Kingdom (UK), the United States (US) and Argentina. Field trials were described for the UK and Argentina, whereas results of a survey conducted in the US have been summarized. Detailed descriptions of materials, mix designs, equations relating lives with material properties and specifications were provided. The authors have stressed the use of two different performance “lives” of porous asphalt mixes – spray reduction life and structural life. Bolzan et al mention that in the UK porous asphalts have been improved with the use of larger size aggregates, use of modified binders, gap-graded aggregate structures and larger air void contents. With respect to use in the US, they mention that European type mixes are gaining much wider acceptance, and that several states have been successful in adopting European concepts and preparing effective specifications. On the Argentinean experience, Bolzan et al provides information on several trials and resultant specifications on the design of porous asphalt mixes. Bolzan et al concludes that for good design of porous asphalt mixes it is necessary to have a high binder content, with modification to prevent draindown, gap-graded aggregate structure, good compaction, use of durable aggregates, high air voids, good construction practices and an effective quality control/assurance system. They recommend that further research should be carried out to make use of the Superpave binder selection method, and to develop maintenance and rehabilitation techniques for extending the spray reduction life of porous asphalt pavements. 1.28.2 Benefits of Permeable Asphalt Mixtures Bolzan et al mentions that the main reason for using porous asphalt is to enhance the driving environment, specifically by reducing noise in all conditions and spray in the wet condition. When used as a wearing course, Bolzan et al indicates that other benefits of permeable asphalt mixtures include reduction of noise in both wet and dry conditions, tire splash and spray in wet conditions, reflected glare at night in wet conditions and

121 reduction in rolling resistance leading to improved fuel economy. They also mention that safety is improved through better tire-road grip, because there is less water on the road surface, increased capacity of roads in wet conditions by enhanced driver visibility, and because of greater texture depth of the road surface under dry conditions. 1.28.3 Materials and Design Bolzan et al mention that there are records of 28 roads trials (with 83 sections) conducted between 1967 and 1991 in the UK (by the Transportation Research Laboratories, TRL). These trials have included a range of materials with aggregates having nominal maximum sizes of 20 mm, 14 mm and 10 mm, various grades of unmodified asphalt binders, binder contents between 3.2 % and 5.7 % and different polymer, fiber and other modifiers. Bolzan et al contend that the durability of porous asphalt can be expressed in terms of the time for which the material effectively reduces spray during wet weather (the spray- reducing life) (or noise during all weather conditions) or the period over which it retains structural integrity (the ultimate life). Based on this contention, the authors have used actual lives or estimated service lives of the different sections to develop performance life versus factors models. Although wide variations have been observed in both spray reduction and ultimate lives, and although site conditions and standard of workmanship have been suspected to be important factors in many cases, the authors have managed to develop models to predict ultimate lives on the basis of important factors such as the aggregate source and size; the bitumen grade; the use of, and type of, binder modifier; and the proportions of these component materials. Before presenting their models, the authors note that the ultimate and spray reducing life of the pavements which have been replaced are directly available, whereas for those that are still functioning after 12 year, estimates of remaining and total ultimate and spray reducing life were made. The total number of trial sections reported in this paper was 68 (or 89 if different lanes, for which separate data were available, are treated as separate sections). The authors also mention that estimation of the spray reducing life is a subjective process, and that their data does not include those from more recent commercial trials or for more developed designs. Bolzan et al then provides a description of the progressive reduction in spray reducing and ultimate life or porous pavements. They mention that the ultimate life of porous asphalt in the road trials generally resulted from progressive binder hardening until the binder was no longer able to accommodate the strains induced by traffic. Brittle fracture was found to begin during winter and, if the surfacing survived a cold winter, it was usually found to remain serviceable during the following warmer months. They mention that an indication of imminent failure was often provided by the onset of raveling, particularly during the winter, which was usually accompanied by an increase in texture depth. Bolzan et al indicates that at this stage, core samples were taken from the surfacing and the condition of the binder was used to establish a failure criterion. Based on the observation from the field performance and results of tests conducted on binder extracted from in-place cores, equations were developed relating ultimate and

122 spray reducing lives to mix design variables. These equations and inferences are shown in Table 68. Table 68: Performance Life Equations from UK Trials Life Equation Inference Ultimate Lu = -3.3 + 0.5 10-4T + 0.4A + 1.4 10-3P + M + 0.6B, ( Radj² = 0,34 ) Spray reduction Ls = 11 – 0.6 10-4T + 0.2A + 3 10-3P + 1.5M – 2.3B, ( Radj² = 0,38 ) 1. Traffic has only a small effect (beneficial for the ultimate life and detrimental for the spray- reducing life); 2. The use of larger nominal sizes of aggregate and of softer grades of bitumen are both beneficial; 3. The inclusion of modifiers appears to be beneficial over and above the increased binder content they permit; and 4. An increase in binder content extends the ultimate life but reduces the spray-reducing life. Note: LU = ultimate life (years), LS = spray-reducing life (years), T = traffic intensity (cv/l/d), A = nominal size of aggregate (mm), P = penetration of the base bitumen (mm/10), M = modified (1) or unmodified (0) binder, B = proportion of binder in the mixture by mass (%), Radj = correlation coefficient adjusted for the degrees of freedom. Bolzan et al provides a description of different materials used in the UK road trials, and some basis for the selection of such materials. These descriptions are shown in Table 69. Based on the field observations, the authors make the following conclusions: 1. For aggregates, 20 mm porous asphalt grading, particularly when using modified binders at higher binder contents, is an effective compromise that maximizes durability at the expense of some loss of hydraulic conductivity, and hence spray- reducing life. 2. The use of high penetration graded asphalt binders enhances durability, but at the expense of earlier closing-up of the surfacing. The extra durability is related to the time taken for the binder to harden to the critical condition, when it can no longer accommodate the traffic induced strains at low temperatures.

123 Table 69: Use of Different Materials in the UK Road Trials Material Basis of use Specification/Type used Observation from field trials Aggregate 1. Tire-induced stresses are applied to relatively few point-to-point contact areas between the essentially single-sized coarse aggregate skeleton; and, 2. relatively cubicle aggregates will provide good drainage and enhance the potential spray-reducing life. UK Specification for Highway Works requires the coarse aggregate in porous asphalt to have: 1. A minimum 10 per cent fines value of 180 kN, 2. A maximum aggregate abrasion value of 12; 3. A maximum flakiness index of 25, 4. A minimum polished stone value dependent on the design traffic intensity, 5. A maximum aggregate impact value of 30 %; and 6. A minimum magnesium sulphate soundness value of 75. Aggregates used were strong aggregates with low flakiness index. The aggregates in porous asphalt are gap-graded such that they consist of coarse aggregate bound with a fine mortar. The proportion of aggregate in the smaller fractions needs to be restricted in order to avoid choking the porous asphalt and, hence, reducing the hydraulic conductivity. For 20 mm porous asphalt, the coarse aggregate is in the 20 mm to 14 mm fraction and the gap should be in the 10 mm to 6.3 mm fraction. The average ultimate life of porous asphalt was: 5 years (with a range of 1 to 12 years) for 10 mm aggregate mixtures and 8 years (with a range 0 to 15 years) for 20 mm aggregate mixtures, The spray-reducing life of porous asphalt, which is not unrelated to the effectiveness of the surface in reducing noise, was 4 years (with a range 2½ to 6½ years) for 10 mm aggregate mixtures and 6½ years (with a range 2 to 8 years) for 20 mm aggregate mixtures.

124 Table 69 Continued: Use of Different Materials in the UK Road Trials. Material Basis of use Specification/Type used Observation from field trials Asphalt Binder High binder contents improved the durability of porous asphalt by providing a thicker binder film, but they also reduced permeability by filling the pores. The maximum binder content is limited by the tendency of excess binder to drain from the mixture, which can result in areas of the finished mat which are either binder-rich or lean and lacking in fines; the binder-rich areas will have inadequate permeability while the binder-lean areas may be prone to premature raveling. The test was first used on a series of trials in 1987 that showed that the most satisfactory target binder content, around 4.5 % for 20 mm nominal size aggregates, was difficult to achieve with unmodified binder without draindown. The penetration of the binder from the sections of porous asphalt in UK trials was monitored using samples taken from the mixing plant tanks and after recovery from the surfacings when laid together with recovered binder from cores taken at various intervals during the life of the surfacings. The results of binder recoveries are subject to several sources of error such as variation within the surfacings (sampling errors) and errors introduced by dissolution, recovery and testing. The errors are particularly evident for polymer-modified binders when the polymer is less than 100 % soluble in the solvent used to extract it. Also, there was evidence of polymer instability in sections from both trials. 1) There was general hardening with time; between mixing and laying, a typical reduction in penetration of 30 % was observed, although there were wide variations; 2) The hardening proceeded at about 20 % reduction in penetration per year; 3) Irrespective of the presence or type of modifier, the critical binder penetration was judged to be about 15 mm with the softening point generally close to 70 °C, after which failure generally occurred when sub-zero temperatures were next encountered; 4) The higher binder content materials showed a slightly lower hardening rate; 5) The presence of hydrated lime tended to lower the hardening rate; and 6) Binders with EVA tended to behave as harder binders with a consequential reduction in longevity. Predictions based on those findings apply to unmodified binders with binder contents of about 4 %(less than the more ideal target binder content of 4,5 % discussed above). The predicted ultimate life of porous asphalt with 100 pen bitumen is 7 to 8 years and that with 200 pen bitumen is 10 to 11 years. However, 100 pen bitumen is still preferable to 200 pen unmodified bitumen for heavily trafficked roads in order to minimize early closing up of the porous asphalt and hence loss of the desired spray- and noise reducing properties. Modifier Modifiers may be used to increase the binder content that can be incorporated into the mixture without the occurrence of binder draindown. These modifiers include both polymer-modifiers and fibers, which increase the surface area over which the bitumen will be spread. Polymer-modifiers can be regarded as part of the binder whilst fibers modify the mixture rather than the bitumen. Polymer modifiers that have been tried include natural rubber, styrene- butadiene-styrene (SBS) block copolymer, ethylene vinyl acetate (EVA), epoxy resin and hydrated lime. ---

125 In describing current practice of using permeable asphalt mixes in the UK, Bolzan et al mention that the UK has finally accepted that permeable asphalt mixes can be durable, but has not pursued many projects with this mix. This is because the thin surfacings have gained a wider acceptance at this time, on the basis that they provide similar benefits as porous mixes at reduced cost and with better durability. This decision has also been affected by problems and need for replacement of porous mixes in two prominent projects. Bolzan et al then provide a summary of the use of porous asphalt mixes in the US, with some insight into recent developments. Their descriptions are based on the results of a survey (published in a circular) conducted by the TRB Committee on Characteristics of Bituminous-Aggregate Combinations to Meet Surface Requirements, A2D03, in 1998. For the convenience of this review, the results, as shown by Bolzan, are summarized in Table 70. Table 70: Summary of results from a Survey on the Use of Open-Graded Friction Course (OGFC), as Presented by Bolzan et al (based on response from forty-two states) Topic Results of survey States using OGFC Nineteen indicate that they use OGFC mixtures. Some agencies construct more than a thousand lane-km per year, others only a few. Seven agencies construct more than 300 lane-km per year. Another ten agencies routinely construct some open-graded mixtures each year. The remaining agencies have either discontinued use or have never used open-graded mixtures. Materials and mixes used in states using large quantities of OGFC mixes Florida uses only one gradation for OGFC. A modified binder is used that is composed of an AC30 bitumen with 12 % (by weight of binder) ground tire rubber. Several aggregates are allowed including crushed granite, blast furnace slag, crushed oolitic limestone (high friction limestone) and lightweight aggregate. Arizona DOT specifies two different OGFC mixtures, one for unmodified binder and the other for rubber modified binder. The bitumen for the unmodified mixture is PG 64-16. For the modified mixture, the base bitumen is PG 64-16 except in colder locations (at high altitude) where PG 58-22 is used. The binder is bitumen modified by the addition of 20 % (by weight of binder) of ground tire rubber. Arizona uses several criteria to specify acceptable aggregate including proportion of carbonate, crushed faces, flakiness index, Los Angeles abrasion, sand equivalent, water absorption and combined bulk specific gravity. Binder content is determined by a formula that depends upon aggregate water absorption, aggregate specific gravity and proportion passing the 2.36 mm sieve. A binder drainage test is not required for the rubber-modified porous asphalt mixture because the binder is very resistant to binder drainage.

126 In showing examples of recent changes occurring to open-graded mixes in the US, Bolzan et al mention the following: 1. Modified binders are being used to reduce the potential of asphalt binder draindown during construction. They mention that modified binders also produce more durable mixes with less aging and potential for raveling. 2. Larger aggregates, with nominal maximum size of 12.5 mm and 16 mm are being used instead of previously used 9.5 mm, to provide larger voids, and hence reduce the chance of clogging. In citing examples of use of this “new generation” type OGFC mixes, Bolzan et al provide a description of practices of the Georgia Department of Transportation (DOT) which was one of the first to adopt these new mixes. Georgia DOT distinguishes between the older and the new mix designs by labeling them as Open-Graded Friction Course (OGFC) and Porous European mix (PEM), respectively. In describing the Georgia DOT practice, Bolzan et al dwell primarily on the use of modified binders and their benefits. They point out the following benefits of using modified binders. 1. Less susceptibility of draindown during construction and service. Draindown which is the separation of the binder/fine aggregate mastic from the coarse skeleton, can occur in a mixture storage silo or in the truck during transport. They mention that mixtures that have suffered from binder drainage produce binder rich areas on the road that have a flushed surface and no voids as well as other areas with little binder and high voids that quickly ravel. 2. Retention of thicker binder films. They mention that thick films of unmodified binder tend to drain downward with time during hot summer weather. The remaining thinned films on the surface particles age and become brittle more rapidly. When the binder becomes sufficiently brittle, aggregate particles are dislodged by traffic and the layer ravels. Modified binders retain film thickness, thereby, reducing aging and stone loss. Bolzan et al also mention that PEMs have higher air voids than OGFC mixtures; this is important since continued benefit from an open-graded mixture is dependent on the void structure remaining open. If the voids are clogged with road debris and winter sands, the effectiveness of the OGFC mixes in draining water is reduced. Bolzan et al mentions that increasing the voids to 20 percent or more provides more resistance to clogging, larger voids tend to be cleaned by hydraulic action of traffic during rainfall, particularly on high-speed pavements. The authors mention that Georgia DOT requires PEM mixtures to have polymer- modified binder and fiber stabilizing additives. PG 76-22 bitumen is typically modified with an SBS or linked SB polymer. Polish resistant, crushed aggregate is required. The mix design criteria include binder content, retained coating after boiling and resistance to binder draindown.

127 Bolzan et al provide a comparison of porous mixes used in the US and in Europe, on the basis of gradation, air voids, aggregates and binders. This comparison has been summarized in Table 71. Table 71: Comparison of US Mixtures to European Mixtures Mix design/material European practice US practice Gradation European gradations allow for a more gap-graded mixture than North American mixtures, although not always The Georgia specification for Porous European Mix is similar to the gradation specified in South Africa. Air voids All European agencies specify minimum air void contents Only one US agency specifies minimum air void content; some US agencies do not compact specimens at all; Air void contents of the US mixtures tend to be considerably lower than European mixtures; Georgia DOT, who developed a specification patterned after the European approach, found permeability of their new mixture to be more than double that of conventional OGFC. Aggregate European agencies generally demand higher standards for aggregates than do US agencies. Los Angeles abrasion values are specified from 12 to 21 %. For open-graded mixtures, US agencies specify 35 to 40 % Binders European agencies use modified binders almost exclusively US agencies are shifting toward the use of modified binders. In the next sections, Bolzan et al provided detailed descriptions of the use of porous asphalt mixes in Argentina. They present results of field trials, general experiences and conclusions and recommendations based on these experiences. Bolzan et al mention that the use of porous asphalt in Argentina started as a result of a search for improvement of wet friction properties and reduction of the potential for hydroplaning. The first use was on a toll road in east central Argentina, connecting

128 Buenos Aires with Mar del Plata. The design was based on European and US experience, with consideration of Spanish, Dutch and French specifications. Principal adoptions were the use of relatively large maximum aggregate size, larger voids and the use of modified binders. Details of the field trials and the experiences, as reported by Bolzan et al, are summarized in Table 72. Table 72: Description of Full Scale Field Trials of Porous Asphalt Mixes in Argentina Topic Description Quality control Brookfield Rotational Viscometer (ASTM D 4402) and torsional elastic recovery test (Spanish Standard NLT 329) were adopted as quality control (QC) tools for the modified binder on site. The Spanish Cantabro Test (NLT 352) was employed as a mixture QC parameter. In-place hydraulic conductivity was measured during construction; a special specification was written based on the European experience and the full-scale trials. In the field, a detailed QC/QA plan was implemented. Description All trials were conducted on Road No. 2, which is a 400 km-length road that has two carriageways of two lanes each. The rainfall in the region is around 800 to 1000 mm per year; road has hard shoulders although it is not a highway because it has numerous crossings along it from the surrounding rural areas. These intersections are an inconvenience for the porous asphalt because many of the vehicles from the feeder roads carry soil that can clog the voids in the material. Time of construction The first full-scale trial was laid down in 1997 on a 300 m-length section with two layers of porous asphalt, one placed as the base layer and the second as the wearing course. The success of the trial was followed by the construction of an entire 22 km- length section (2 lanes on both carriageways). A year later, a 100 km section was constructed divided into four sections. A total of more than 100,000 tonnes were placed in this project in about two years. Following this project, other roads within Buenos Aires Province followed the trend giving a total of more than 170,000 tonnes in a little over two years. Binders The modified binders used were of two types, EVA-modified and SBS-modified, with different levels of modification that, in terms of torsional elastic recovery at 25 °C, ranged between 35 and 75 % of its original shape; Two levels of modification, based on the torsional elastic recovery test, were set at not less than 40 % and 70 %. Aggregate Aggregates comprised of granite from three locations in the south-west of Buenos Aires Province: Olavarria, Tandiland, Balcarce Structure The layer thickness in all cases was 40 mm and the maximum particle size was mostly 19 mm. Other trials have been undertaken in order to compare binder types, different aggregate gradations and two thicknesses (40 and 50 mm). A twin-layer porous asphalt was constructed along one kilometre length using two different formulas. In this twin- layer system, a 20 mm thick layer with 12 mm maximum aggregate size was placed on top of a 40 mm thick layer with 19 mm maximum aggregate. Both layers were design and controlled separately.

129 Table 72 Continued: Description of Full Scale Field Trials of Porous Asphalt Mixes in Argentina. Topic Description Mix Design The Cantabro test (NLT 352/86) was adopted in the two modes: dry (6 hours at 25 °C, 300 revolutions) and wet (24 h at 60 °C, test at 25 °C, 300 revolutions); the binder drainage test was adopted from the UK; first trial was conducted according to the Spanish specifications, with two aggregate gradation bands, a modified binder with up to 15 % elastic recovery, Portland cement as a filler, a 12 mm aggregate maximum size and an EVA modified binder. Later trials were conducted with SBS modified binders, hydrated lime as filler (2.5 to 3 %) and a 19 mm aggregate maximum size. The layer thickness was to be 40 mm. The mixture was composed of 86 % coarse aggregates (6 to 20 mm), 11 % fine aggregates (0 to 6 mm) and 3 % hydrated lime. The binder content was established at 4.5 % of the total weight of the mixture by performing both Wet and Dry Cantabro tests and the binder drainage test. The total voids content in the mixture was set at 23 %. The in-situ total air voids content ranged from 18 to 25 %. In the case of the 12 mm maximum aggregate size mixture, the composition was 80 % coarse aggregate (6 to 12 mm), 17 % fine aggregate (0 to 6 mm) and 3 % hydrated lime. The binder content was fixed at 5.0 % by weight of the total mixture. The Cantabro (25 °C, 300 revolutions) and Indirect Tensile (25 °C, 50 mm / min) tests were performed on all the mixtures both on plant-prepared Marshall briquettes and on cores taken from the pavement. The Cantabro test values at 25 °C were in the range of 10 to 17 % for plant produced mixture and between 15 and 20 % for the cores. The acceptance criteria were maximum values of 25 % in the dry condition and 35 % in the wet condition test. The wet condition test was conducted every 3,000 tonnes of plant produced material with the results showing an average 22 % abrasion loss. The Indirect Tensile test was performed mainly as a mean to verify the uniformity of the cores extracted from the road. Thus, every core was subjected to the Indirect Tensile test and the results showed a resistance in the range of 0.5 to 0.6 MPa. Prior to testing, the specimens were analyzed for voids content, thickness, density and exterior appearance. Aggregate Los Angeles Abrasion test values were between 20 and 23 % for gradation B (ASTM C 131). No material was allowed with Los Angeles values higher than 25%. All the particles were 100 % crushed, with a Flakiness Index not greater than 25 %, and were binder compatible (there was no need of an anti-stripping additive). The finer gradation had a Sand Equivalent Test value of 70 %. Accelerated Polishing Stone Value higher than 45 were adopted from the Spanish recommendations, although the British require values nearing 60. However, local granites can only reach 48 to 50 PSV. Gradation The first trial was performed using a gradation envelope following the Spanish PA12 gradation band. Later it was adjusted to a different gradation. ------------------------------------------------------------------------------- PA12 Sieve mm 19.0 12.5 9.5 4.75 2.36 0.63 0.075 Percent Passing 100 70-100 50-80 15-30 10-22 6-13 3-6 Revised Percent Passing 100 70-90 50-80 15-20 10-18 6-13 3-5 -------------------------------------------------------------------------------- For the mixture to be efficient at allowing water to drain through it the proportion passing the No.4 sieve was kept in the range of 15 to 19 % and the difference between sieves No.4 and No. 8 was kept at less than 8 %. The material passing 74 microns was kept at 4 %, and a 19 mm maximum size was adopted in two out of the three projects.

130 Table 72 Continued: Description of Full Scale Field Trials of Porous Asphalt Mixes in Argentina Permeability Of the different types of available falling head permeameters, There are several types and Argentina adopted the LCS type, standardized by the Spanish in NLT-327/88, for measurement of hydraulic conductivity. This piece of equipment works by percolating water through the material while taking two readings as the water goes down. Once the reading in seconds is obtained, it can be converted into the air voids content in the mixture through the following equations: Ln (K) = 7.624 – 1.348 Ln (T), Ln (H) = 4.071 – 0.305 Ln (T) where: K = the permeability coefficient (cm/s 10²), H = the proportion of air voids in the mixture (%), T = the evacuation time (s); correlation was developed on the basis of mixes containing between 3.5 and 5.5 % modified binder, maximum aggregate size of 10 to 12 mm, between 10 and 15 % of aggregate passing the No.8 sieve, having between 2 and 6 % of aggregate passing the No.80 sieve. The total air voids content measured in the projects ranged from 18 % to 25 % with the mean value between 21 and 23 %. Bolzan et al mention that based on the first project and European experience, a special specification was developed around mid 1998. Early in 1999, before the start of the HighwayEzeiza-Cañuelas (near the international airport) project, a modified version of the original specification was made that incorporated recent experience, more user- friendly modified-binder specifications and tighter tolerances. The authors mention that the tighter tolerances were on the binder content, the aggregate gradations, the volumetric properties and the filler content variations during construction. Also, additional requirements were placed on the construction of the asphalt base course because this layer has to have enhanced properties of in-place voids content, fatigue resistance properties and rut resistance characteristics. Bolzan et al discuss some structural considerations of porous asphalt mix layers in their paper. On the basis of the facts that porous asphalt is a mixture in which fractions of the aggregate grading are absent, they contend that porous asphalt mixtures have lower strength than dense-graded mixtures. They mention that some researchers accept that these mixtures have up to 70 % of the strength of a conventional mixture; others indicate the ratio is only 50 % whilst the Spanish believe that they are structurally equivalent with conventional dense-graded asphalt mixtures. They are also considered to be less shear stress resistant. Bolzan et al indicates that Argentina adopted a 50 % structural capacity for porous asphalt mixtures in the initial projects. The resilient modulus (ASTM D 4123) at 25 °C and 10 Hz of porous asphalt mixtures, prepared in the laboratory of Argentina, was found to be about 2200 MPa, approximately 60 % of the conventional mixtures. However, Bolzan et al point out that at both higher and lower temperatures, polymer- modified porous asphalt mixes perform better than unmodified conventional mixes and that further research needs to be conducted to reach at definitive conclusions. In discussing the quality control and quality assurance aspects in construction of porous asphalt mix courses, Bolzan et al mention that a QC/QA system was developed and implemented. The system comprised of control testing by the contractor, quality assurance testing by the owner; and independent assurance sampling and testing to measure resilient modulus of the mixture and initial friction properties. Plant controls and

131 checks consisted of all the operations related to the control of the quality of the incoming materials (binder, aggregates, filler, etc), periodic controls during the construction stage and testing of the end product. The key components of the program were: 1. Volumetric Quality Control, 2. Materials Quality Certification, 3. Technical Evaluation through Data Analysis; and 4. Acceptance based on end product specification. Bolzan et al indicates that the specifications are based on Marshall density, relative hydraulic conductivity, total in-situ air voids content, binder content, surface tolerance and texture depth. The main goal is to obtain sufficient interconnected air voids in order to provide enough relative hydraulic conductivity and to maintain a high pavement surface texture depth. Non-destructive testing employing two Falling Weight Deflectometers (FWD) were carried out to evaluate the pavement structural capacity achieved as part of the QC/QA system. In the Ezeiza-Cañuelas Highway project, a Kuab FWD was used for the first 24 km and a Dynatest FWD for the second 12 km length. A special testing protocol was written specifying measurements on the outside wheel path of the four lanes at 50 and 100 m intervals applying a 40 kN load. Based on empirical measurements and the estimated future traffic, a maximum initial deflection of 500 microns was set. All the individual maximum deflection values measured were lower than the specified value. In discussing durability expectations, Bolzan et al points out that the spray or the functional life ends with the clogging of interconnected voids, whereas the ultimate or structural life ends with raveling. The use of polymer modified binder and hydrated lime were made to optimize the aggregate-binder bond. They mention that the functional life has been estimated as being between 3 and 4 years, and the structural life has been assumed to be 8 years. 1.28.4 Construction Practices No information on construction practices has been presented. 1.28.5 Maintenance Practices No information is provided on maintenance practices of friction course. 1.28.6 Rehabilitation Practices No information is provided on rehabilitation practices of friction course. 1.28.7 Performance Relevant information on performance has been presented in the mix design section. 1.28.8 Structural Design Relevant information on structural design has been presented in the mix design section. 1.28.9 Limitations Although not labeled as limitations, Bolzan et al provide several disadvantages of using porous asphalts in their opening paragraph. They mention that these disadvantages include increased costs, relatively low structural strength, due to its high void content,