Influence of Reclaimed Asphalt Pavement (RAP) on Surface Friction

Transcription

1 0 0 Influence of Reclaimed Asphalt Pavement (RAP) on Surface Friction Karol J. Kowalski Assistant Professor and Scholar of Foundation for Polish Science Warsaw University of Technology Faculty of Civil Engineering, Office 0 Al. Armii Ludowej Warsaw, Poland, 00- tel. + () fax. + () k.kowalski@il.pw.edu.pl Rebecca S. McDaniel Technical Director North Central Superpave Center (NCSC) Montgomery Street, P. O. Box West Lafayette, IN 0 tel. () - x fax. () -0 rsmcdani@purdue.edu Jan Olek Professor and Director, NCSC Purdue University School of Civil Engineering, Office G 0 Stadium Mall Drive West Lafayette, IN 0 tel. () -0 fax. () - olek@purdue.edu Date of re-submission: November 0 Word count: 0 Figures (0): 0 Tables (): TOTAL: 0 TRB 0 Annual Meeting CD-ROM

2 Kowalski, McDaniel and Olek ABSTRACT Reclaimed asphalt pavement (RAP) is currently a widely used material for the construction of asphalt pavements. However, in regions deficient in non-polishing aggregates, RAP is not commonly allowed in mainline surface courses for high volume roadways because of concerns about frictional performance. The goal of the present study was to determine the maximum amount of RAP that can be blended with high friction aggregates and used in surface mixes without significantly impacting their frictional properties. The initial part of the study described here included a comparison of RAPs collected from six different sources (mixing plant stockpiles) in Indiana. It was shown that the field-collected RAP s exhibited fairly consistent properties in terms of their gradations and binder contents. In the second part of the study, low friction aggregate (limestone) was used to produce a worst case scenario RAP for evaluation of its influence on the frictional characteristics of two types of hot mix asphalt mixtures: (a) dense graded asphalt (DGA), and (b) stone matrix asphalt (SMA). The DGA and SMA mixtures were produced with various amounts of this laboratoryproduced RAP. The RAP was blended with two types of highly friction resistant aggregates: steel slag and air cooled blast furnace slag. Overall, the results suggest that for the materials and mixtures studied, the maximum amount (threshold level) of RAP that can be used in surface mixes without detrimental effect on their frictional properties was about 0% even when the RAP contained a highly polishable aggregate. Key words: Reclaimed Asphalt Pavement (RAP), friction, Stone Matrix Asphalt (SMA), dense graded asphalt (DGA), asphalt pavement, laboratory mix evaluation, Superpave INTRODUCTION Reclaimed asphalt pavement (RAP) has been used for many years in the USA and elsewhere in the production of hot mix asphalt (HMA). When properly designed and constructed, pavements including recycled asphalt can perform as well as or better than pavements constructed from virgin materials. The use of RAP in pavements is desirable since it offers economic benefits without compromising performance. From the sustainability point of view, recycling reuses the existing aggregates and RAP binder, thus reducing the need for new materials. In addition, recycling can reduce transportation costs and expenses associated with landfilling or storing of the milled material. There are also additional environmental benefits of reusing existing resources that are difficult to quantify. While, in general, the Departments of Transportation (DOT s) already make extensive use of RAP, there are still some applications where this material has not been used to full advantage. Historically, some DOTs have not allowed the use of RAP in pavement surface courses due to concerns about potentially negative effects on pavement friction. Since it is difficult to know specifically what types of aggregate are present in RAP, their effects on friction are unknown. This is especially a concern in regions with predominantly soft aggregates (e.g. limestones) which can be susceptible to polishing. Under the current economy, there is an increased interest in using higher amounts of RAP in more applications. As a result, some states are considering expanding and revising their specifications regarding the RAP usage. Recently, for example, the Indiana DOT began to allow the use of RAP in surface mixes, both Superpave dense-graded and Stone Matrix Asphalt (SMA), at up to % by weight of the total mixture (). TRB 0 Annual Meeting CD-ROM

3 Kowalski, McDaniel and Olek Concerns with using RAP in Surface Mixtures There are two potential problems with using RAP in surface mixtures. Friction resistance is the primary concern in many states. When the aggregates present in the RAP are of unknown origin, their potential effects on friction are of concern. A secondary concern is the possibility that using too much RAP (or using RAP that is too hard) could over-stiffen the surface course, making it more susceptible to cracking or to raveling. Both concerns are highly related to the fact that RAP is typically removed from old roadways and may contain different types of aggregates, binders, patches, chip seals, etc., all intermingled in one stockpile. When properly processed and stored, however, the variability of RAP can be controlled and may not be as high as perceived (). Proper processing and storing of RAP has been extensively discussed in various publications of the National Asphalt Pavement Association (,). Pavement Friction Pavement friction is primarily a function of the surface texture, which includes both micro- and macrotexture. Overall texture is affected by the aggregate shape as well as by the aggregate size and gradation. Pavement microtexture can only be modified by changing the types of aggregates used in the mixture while macrotexture can be modified by changing the size and gradation of the aggregates (,). In order to produce a blend which can improve pavement frictional resistance, aggregates with high frictional characteristics may be combined with low friction aggregates (,). The influence of surface texture on the frictional characteristics of pavements can be represented by their polishing rate and terminal friction value (minimum, steady friction value achieved after polishing). A model allowing for determination of these values (based on changes in friction and macrotexture) was published elsewhere (). Once established, the terminal friction value can be compared with other parameters, such as the International Friction Index (IFI) flag value, and can be used for the evaluation of the frictional properties of mixes containing RAP. The flag value is defined as the friction number at or below which a site investigation needs to be conducted (0). This approach was previously proposed by Kowalski et al. (). PROBLEM STATEMENT AND OBJECTIVES The main purpose of this research was to determine if it may be acceptable to use RAP in mainline surface courses of high volume roadways without significantly decreasing the pavement s friction properties. Additionally, the goal was to determine the maximum amount (threshold level) of RAP, regardless of the type of aggregate it contained, which could be used in surface mixes without detrimental effect on their frictional properties (assuming such a threshold exists). TEST PROGRAM The program of this study can be divided into the following seven steps:. RAP samples from six different locations across the state of Indiana were collected and extracted in order to determine their binder content and aggregate gradation and to assess the overall variability of RAP.. Using historical data and guidance from the Indiana DOT Office of Materials Management, a highly polishable aggregate exhibiting low frictional properties was selected. This aggregate was a limestone coarse aggregate, and it is referred to herein as a poor quality aggregate in TRB 0 Annual Meeting CD-ROM

4 Kowalski, McDaniel and Olek 0 0 terms of its frictional properties. This aggregate would not be allowed for use in moderate or high traffic volume applications (). Using the materials and information obtained in step (RAP composition), a mix was produced in the laboratory and aged in order to fabricate a worst case scenario RAP, in terms of frictional properties. This RAP will be referred to in this paper as the laboratory produced RAP.. Design Mix Formulae (DMFs) for Superpave Hot Mix Asphalt (referred to here as a Dense Graded Asphalt, DGA) and Stone Matrix Asphalt (SMA) mixes were obtained to serve as examples of typical Indiana surface mixes.. Using information from step, aggregates and binders commonly used in DGA and SMA mixtures for high volume roads in Indiana were collected. Using those materials and the laboratory produced RAP, mixes with various RAP contents were designed.. SMA and DGA mixes, designed in step, were fabricated in the laboratory and compacted in order to produce specimens for friction testing.. Specimens were polished in the laboratory using a specially designed Circular Track Polishing Machine (CTPM). In addition, texture and friction were measured on the specimens before polishing and at various points during the polishing process to calculate an International Friction Index (IFI) and to evaluate the frictional resistance of the mixes.. Combining the obtained test results with the previously mentioned friction flag value and with additional information collected during the literature review, a recommended RAP threshold level was proposed. Materials and Mix Design The compositions of the actual RAP samples, the laboratory fabricated RAP, and the DGA and SMA mixtures incorporating the laboratory fabricated RAP are described below. RAP Sample Composition RAP samples collected from various stockpiles were extracted following ASTM D, test method B (AASHTO designation T), to determine the aggregate gradation and binder content. The RAP stockpiles consisted of material milled off of various state and local contracts, which was then crushed and screened through a mm sieve. The RAP aggregate gradations, determined according to AASHTO T0, are shown in FIGURE. Analysis of the data in FIGURE indicates that the gradations were fairly similar and consistent, even though the RAP samples were collected from widely dispersed HMA plants across the state of Indiana. The greatest variation in the gradation occurred on the. mm (No. ) sieve, where the maximum difference between the various RAPs was equal to %. It could also be noticed that gradation of RAP_ is slightly different than the gradations for other five RAPs. The binder content for RAP_ is equal to.% and is also a little higher than the binder contents for the other samples (which were between.% and.% with an average of.%). TRB 0 Annual Meeting CD-ROM

5 Kowalski, McDaniel and Olek 00 0 RAP _ Cumulative % Passing 0 0 RAP _ RAP _ RAP _ RAP _ Ma ximum De ns ity Line 0 0 RAP _ Sieve size, mm FIGURE Gradation of six RAP samples. Laboratory Produced RAP Based on the analysis of the compositions of RAP across the state, an average RAP was proposed and fabricated in the laboratory. The gradation of the laboratory produced RAP is shown in FIGURE. This laboratory produced RAP contained.% of PG - binder and limestone that polishes substantially when exposed to traffic. This limestone was found to be one of the lowest quality aggregates, in terms of pavement friction, available in Indiana, so it can be assumed that the RAP produced with this aggregate may be considered as representing the worst case scenario. The limestone aggregate was delivered to the laboratory, oven dried at 0 C (ºF) and cooled to room temperature prior to being sieved and sorted into individual size fractions. The aggregate was then batched to produce the desired blends. Prior to mixing, the batched aggregate blends (and the binder) were heated to a mixing temperature of 0ºC (0ºF). The mixing was performed in a five-gallon, bucket type laboratory mixer, which was first primed with a butter mixture in order to avoid binder loss during preparation of the test specimens. Next, the mix was conditioned for two hours at the compaction temperature ( C or ºF) according to AASHTO R 0. After conditioning, the mixture was left in an C (ºF) oven for hours, to simulate the aging that occurs over the service life of a pavement. After this exposure, the mixture was cooled and re-mixed in the laboratory mixer to separate it into particles smaller than. mm. The RAP was then stored in closed buckets until the start of specimen preparation process. TRB 0 Annual Meeting CD-ROM

6 Kowalski, McDaniel and Olek SMA and DGA Mix Designs Four SMA and four DGA mixes, each with four levels of RAP content (0%, %, % and 0%), were designed and tested in this study. The mixes were designed based on the typical Indiana DMFs for.mm Nominal Maximum Aggregate Size (NMAS) mixes. The target gradations of the SMA and DGA blends are shown in FIGURE. The laboratory produced RAP was blended with steel slag (SS) for the SMA mixes and with air cooled blast furnace (ACBF) slag for the DGA mixes. (It should be noted that the ACBF slag is not currently allowed for SMA surface mixes in Indiana.) In addition, a mineral filler (lime) as well as cellulose fibers (in the amount of 0.% by weight) were also added to the SMA. 00 Cumulative % Passing "La bora tory P roduc e d" RAP DGA Control P oint DGA: Ta rge t Gra dation Ma ximum De ns ity Line S MA: Ta rge t Gra da tion S MA Control P oint Sieve size, mm FIGURE Gradations of SMA and DGA mixtures and of laboratory produced RAP. A PG - binder was used to produce all mixes. It is important to note that the binder grade was not changed for the entire range for RAP percentages added (i.e., no grade bumping was performed). Also, PG - binder is not the typical grade used for SMA mixtures in Indiana; a (nominally) softer binder was used here in order to reduce the compactive effort needed to fabricate the test specimens for friction measurements. The design binder content for each mix was that which provided % air voids (V a ) in the mix compacted in the Superpave Gyratory Compactor (SGC). The compaction effort used (N design ) was equal to 00 gyrations for the SMA and for the DGA mixtures. This compactive effort corresponds to an anticipated high traffic level (>0 million ESALs (Equivalent Standard Axle Loads)). The design process was conducted following AASHTO M and AASHTO M for the SMA mixes and AASHTO M and AASHTO R for the DGA mixes. Details of the mix design are shown in TABLE. It should be noted that the binder content shown in TABLE includes the binder from the RAP. TRB 0 Annual Meeting CD-ROM

7 Kowalski, McDaniel and Olek 0 TABLE Mixture Type, Material and Volumetric Data HMA Type DGA SMA RAP Content, % (by Weight) RAP Content, % (by Volume) Main Aggregate Type Air Cooled Blast Furnace Slag Steel Slag Main Aggregate Content, % 00 0 Mineral Filler, % Binder Content, P b, % Bulk Spec. Grav., G sb Max. Theor. Spec. Grav., G mm Specimens Preparation and Testing Methods Based on the mix designs described above, eight different mixes (four SMAs and four DGAs) were fabricated in the laboratory. After mixing, they were placed in the oven for two hours to allow for binder absorption, following AASHTO R 0. They were then cooled down and stored in closed buckets until the time when compaction was initiated. In order to determine the frictional properties of the various mixtures, a test procedure developed in another study () was utilized. This procedure is briefly described here. First, slabs are fabricated from the mixture to be tested. Laboratory produced HMAs are reheated to the compaction temperature. Based on the volume of the mold and the specific gravity of the mix, the approximate weight of mix that would yield to % air voids was determined. That amount of mix was then placed in a square wooden mold (0 mm ( in.) by 0 mm ( in.) and mm (. in.) deep) and compacted using a large rolling pin mounted on a fork lift. Once compacted, the slabs were allowed to cool thoroughly. Following the compaction, the specimens were subjected to polishing and measurement of their frictional properties. Polishing was performed using a device called a Circular Track Polishing Machine (CTPM), shown in FIGURE. This device consists of three rubber tires attached to a rotating plate. The wheels travel over the same footprint as that used in the devices to measure friction and texture (described below). The polishing wheels travel at approximately rotations per minutes (RPMs). Since each revolution rotates three tires over the same track on the surface, there are about cumulative wheel passes per minute. Water is sprayed on the slab surface to help remove the debris generated during polishing. During polishing, a total load of 0. kn is applied through the tires to the surface. TRB 0 Annual Meeting CD-ROM

8 Kowalski, McDaniel and Olek 0 0 FIGURE Circular Track Polishing Machine (CPTM). Before the polishing was initiated (and periodically during polishing), the surface texture and friction of the slabs were measured. The surface texture was measured using a laser-based Circular Track Meter (CTM), following ASTM E. The texture is reported in terms of the Mean Profile Depth (MPD) and measured in millimeters. Then, the friction of the surface was measured using a Dynamic Friction Tester (DFT), following ASTM E. In the DFT device, three rubber sliders attached to the disk are accelerated to tangential velocities of up 0 km/h ( mph) and then dropped onto the surface. The torque generated as the disk slows provides an indication of the friction at various speeds. The main value of interest here is the DFT number at km/h ( mph), designated DF. The previously determined MPD value can be combined with the DF value and used to calculate the International Friction Index (IFI) following ASTM E 0. The IFI consists of two parameters: the calibrated wet friction at 0 km/h (F0) and the speed constant of wet pavement friction (S p ). The polisher was stopped periodically during testing so the measurement of friction and texture could be performed. This occurred after the following cumulative numbers (in thousands) of wheel passes:.,.,,, 0,,, and. Typically, for asphalt mixtures the initial friction tends to be low because of the presence of binder film coating the aggregate particles. After the binder film is worn off by traffic, the friction increases. Continued wheel passes tend to cause a decrease in the friction level, and sometimes changes in the texture, as the aggregate particles undergo polishing. Eventually, the friction tends to level off at the socalled terminal friction value. This occurs when embedded aggregates at the surface are polished as much as they will polish and further wheel passes do not cause additional loss of friction. This general trend in friction is observed both in the field and in the lab. Past research work () has shown that terminal friction can usually be obtained in the CTPM after fewer than,000 wheel passes (,000 CTPM revolutions), even for mixtures with high friction aggregates like steel slag. TRB 0 Annual Meeting CD-ROM

9 Kowalski, McDaniel and Olek 0 TEST RESULTS The results of testing of eight slabs show that both macrotexture (expressed by MPD) and dynamic friction (expressed by DF ) changed during the polishing process. Changes in the MPD are shown in FIGURE for the SMA specimens and in FIGURE for the DGA mixture. It can be observed that the macrotexture of the SMA specimens remained relatively constant during testing while macrotexture of the DGA increased significantly. The greatest rate of increase was observed during the first part of the polishing, up to 0,000 wheel passes, after which it stabilized. Overall, the macrotexture of DGA nearly doubled. During other studies (), it was noticed that a similar phenomenon has also been observed in the field; there was a significant macrotexture change for a DGA pavement and a relatively low change for an SMA. It has to be noted, however, that the initial MPD (before initiating polishing) was much higher for the laboratory fabricated specimens than for the field test sections, perhaps indicating a need to improve the specimen compaction method (which will be investigated in future work). It appears the current specimen preparation method may not apply enough compactive effort.. SMA_00 SMA_. SMA_ SMA_0 MPD, mm no. wheel passes, 0^ FIGURE Mean profile depth (MPD) data as a function of no. of wheel passes for SMA specimens. TRB 0 Annual Meeting CD-ROM

10 Kowalski, McDaniel and Olek MPD, mm DGA_00 DGA_ DGA_ DGA_ no. wheel passes, 0^ FIGURE Mean profile depth (MPD) data as a function of no. of wheel passes for DGA specimens. Based on the analysis of the data shown in Figures and, it can be concluded that an increase in the RAP content resulted in an increase in the MPD for the SMA specimens, while the reverse trend was observed for the DGA mixes. Changes in the macrotexture observed during polishing are, most likely, connected to the raveling of the DGA specimens. The CTPM is a relatively aggressive polishing method in which the shearing action of the tires abrades the surface. In addition to the MPD, the DF parameter was also determined after each increment of polishing cycles. These two parameters were used to calculate the calibrated wet friction (F0) values (following the ASTM E 0), as shown below: F0 0 Sp = DF e () S p =. +. MPD () where: DF = wet friction number measured at the speed of km/h, MPD = mean profile depth (mm). When using Equation with the typical range of MPD values (0. mm to. mm) and DF values (0. to 0.), it could be noted that the F0 parameter is highly influenced by the DF. The trend of the plot of DF versus number of wheel passes is typically similar to the plot of F0 versus number of wheel passes (, ). An example of the typical changes in the DF values taking place during polishing is shown in FIGURE for the SMA mixture with 0% RAP content. It is highly possible that the changes in the MPD values (observed especially for the DGA specimens) may somehow mask lowering of the frictional properties of the specimens caused by polishing (that is, a lower DF is counteracted by the higher MPD and results in smaller changes in the F0). TRB 0 Annual Meeting CD-ROM

11 Kowalski, McDaniel and Olek DF no. wheel passes, 0^ FIGURE Dynamic friction (DF ) data for SMA mixture with 0% of RAP. In order to quantify changes in the F0 values taking place during the polishing and to evaluate the frictional properties of the mixture, a polishing model developed in previous research () was used. This model allows for estimation of the terminal friction level (referred to as a F0@X ) and the polishing rate (a ). Note: X represents the number of wheel passes at which the terminal friction level is reached. The model has a general form shown in FIGURE and in Equations - (). y + FIGURE Polishing model (after ref. ). ( x) = a0 + ax + ax ax () 0 < x x 0 y ( x) = a ln( x + x + a () 0 ) x Initial Pavement Decreasing Life Zone Friction Zone F0 < x 0 x y = a ln( x + x0 ) + a = const. () x > x y I a II y x 0 x Friction Stabilization Zone III y x, log no. of passes TRB 0 Annual Meeting CD-ROM

12 Kowalski, McDaniel and Olek 0 The parameters a 0 to a can be found by minimizing the sum of square errors (SSE) assuming that the minimum SSE would result in the model that best fits the measured data: SSE data predicted by the model - measured data () = ( ) The model generated for the tested mixes yielded relatively high coefficient of determination (R ) values with an average of 0.. Of the eight mixes tested, one mix had an R value equal to 0., another one was 0. and the rest of them had R values greater than 0.. In general, the higher (less negative) the a value is, the more resistant the specimen is to polishing. A high F0 value corresponds to high terminal friction value for the pavement. An example of the typical changes in the F0 values taking place during polishing is shown in FIGURE for the SMA mixture with 0% RAP content. On the same figure, the model fitting the data with R =0. is also shown. It can be observed that after about 0,000 wheel passes the changes in the F0 value were relatively small, suggesting that the specimen had reached its terminal friction level. 0.0 data model F0 0.0 R = no. wheel passes, 0^ FIGURE Calibrated wet friction (F0) data and model predicted F0 values for SMA mixture with 0% of RAP. A summary of the distribution of the terminal friction level (F0@X ) and the polishing rate (a ) parameters is shown in FIGURES and 0, respectively. Changes in the terminal friction level shown in FIGURE suggest that the addition of RAP indeed influences the friction, as could be expected. The more RAP material that is added, the lower the friction value becomes. This general trend can be observed for both DGA and SMA mixtures. For the SMA mixtures, the changes in the F0@X values generally decrease linearly, while for the DGA mixtures the F0 drops more between samples with RAP contents of % and % than between 0 and % or to 0%. It has to be noted, however, that even the lowest F0@ X values (observed for the specimens with 0% RAP content and equal to about 0.-0.) are much higher than the flag value determined in another study (). According to that study, the F0 value should be greater than Moreover, those findings correspond well with findings from another study (), in which DGA mixtures with dolomite aggregate and various amounts of steel slag (ranging between 0-0%) were tested. For a mixture with 0% high friction resistance aggregate (steel slag) and 0% lower friction resistance aggregate (dolomite), the F0@X was equal to 0.. TRB 0 Annual Meeting CD-ROM

13 Kowalski, McDaniel and Olek 0.0 DGA SMA X % (by weight) of RAP FIGURE Distribution of friction terminal value (F0@x ). Changes in the polishing rate shown in FIGURE 0 suggest that the addition of poor quality RAP did indeed influence the polishing susceptibility, as could be expected. The general trend observed for both DGA and SMA mixtures shows that the more polishing susceptible RAP material is added, the less resistant to polishing the combined mixture becomes. It can also be noticed that for specimens with % poor quality (laboratory) RAP the polishing rate was about -0.0 (for both DGA and SMA specimens) and that it was similar to the polishing rate of specimens with no RAP. This suggests that up to % RAP has an insignificant influence on the polishing resistance of the mixture, even when the RAP itself is highly polishable. In another study (), changes of steel slag content from 0 to 0% resulted in a parameter changing from to In this study, the lowest values are similar for both SMA and DGA mixtures and equal to about This polishing rate was observed here for mixtures with about 0% RAP. As observed in the previous study (), a polishing rate of up to about -0.0 is not significant. TRB 0 Annual Meeting CD-ROM

14 Kowalski, McDaniel and Olek polishing rate, a DGA SMA % (by weight) of RAP FIGURE 0 Distribution of polishing rate (a ). DISCUSSION AND CONCLUSIONS The initial part of this study included a comparison of plant collected reclaimed asphalt pavement (RAP) from six different sources in Indiana. It was shown that the field collected RAPs exhibited fairly consistent properties in terms of their gradations and binder contents. Then, a laboratory (worst case scenario) RAP was fabricated using a low quality (in terms of its frictional properties) aggregate (limestone). This laboratory RAP was mixed (in varying quantities) with high quality (friction resistant) aggregate to prepare two types of hot mix asphalt mixtures: dense graded asphalt and stone matrix asphalt. The research showed that the addition of up to % of low quality RAP to the high quality aggregate blend did not significantly influence the mixture polishing rate. At RAP contents up to about 0%, the polishing susceptibility was not severe and it only increased at higher (up to 0%) RAP levels. It was also shown that the terminal friction level decreased when the poor quality RAP content increased. In the mixes with up to 0% RAP, the terminal friction level was much greater than the flag value determined in other study (). Based on the overall results of this study, it seems that the maximum amount (threshold level) of RAP which can be used in surface mixes without detrimental effect on the frictional properties of the surface, is about 0%. It has to be noted, however, that the effects of the RAP binder on ultimate pavement performance (especially susceptibility to cracking) were not studied here. The focus of this research was on the frictional properties only. DISCLAIMER AND ACKNOWLEDGEMENTS This work was supported by the Joint Transportation Research Program administered by the Indiana Department of Transportation and Purdue University. The contents of this paper reflect the views of the authors, who are responsible for the facts and the accuracy of the data presented herein, and do not necessarily reflect the official views or policies of the Federal Highway TRB 0 Annual Meeting CD-ROM

15 Kowalski, McDaniel and Olek Administration and the Indiana Department of Transportation, nor do the contents constitute a standard, specification, or regulation. The authors wish to express their gratitude to Mr. Yu-Min Su, Dr. Ayesha Shah, Dr. Eyal Levenberg and Mr. Wubeshet Woldemariam of Purdue University for their assistance during testing and for the in-depth discussions. Mr. Andrew Szabat s review of the draft manuscript is also appreciated. REFERENCES. Indiana Department of Transportation. 0 Standard Specifications Book. Section 00, Asphalt Pavements, Indianapolis, 0.. Nady, R. M. The Quality of Random RAP: Separating Fact from Supposition. National Asphalt Pavement Association, HMAT magazine, vol., issue, Lanham, MD,, pp. -.. Recycling Hot Mix Asphalt Pavements. Information Series (IS-). National Asphalt Pavement Association, Lanham, MD, 0, pp... Newcomb, D. E., E. R. Brown, and J. A. Epps. Designing HMA Mixtures with High RAP Content. A Practical Guide. Quality Improvement Series, National Asphalt Pavement Association, Lanham, MD, March 0, pp... McDaniel, R. S., and B. J. Coree. Identification of Laboratory Techniques to Optimize Superpave HMA Surface Friction Characterization. Phase I: Draft Final Report. Institute for Safe, Quite and Durable Highways, SQDH 0, West Lafayette, Indiana, 0.. Gardziejczyk, W. Texture of Road Surfaces Methods of Measurement Parameters Evaluation and its Influence on the Tire/Road Noise. Drogi i Mosty, no., Poland, 0, pp. -.. West, T. R., J. C. Choi, D. W. Bruner, H. J. Park, and K. H. Cho. Evaluation of Dolomite and Related Aggregates Used in Bituminous Overlays for Indiana Pavements. Transportation Research Record, vol., 0, pp. -.. Gee, K. W. Surface Texture for Asphalt and Concrete Pavements. FHWA Technical Advisory, Classification Code: T 00., June, 0.. Kowalski, K. J., R. S. McDaniel, and J. Olek. Development of a Laboratory Procedure to Evaluate the Influence of Aggregate Type and Mixture Proportions on the Frictional Characteristics of Flexible Pavements. Journal of the Association of Asphalt Paving Technologists, Vol., pp. -0, Li, S., Noureldin, S. and Zhu, K. Upgrading the INDOT Pavement Friction Testing Program, Final Report FHWA/IN/JTRP-0/, West Lafayette, Indiana, October 0.. Kowalski, K. J., R. S. McDaniel, J. Olek, A. Shah, and S. Li. Development of the International Friction Index Flag Value. 0 th International Conference on Application of Advanced Technologies in Transportation, Athens, Greece 0, Paper No. 0.. Kowalski K. J., R. S. McDaniel, A. Shah, and J. Olek. Long Term Monitoring of the Noise and Frictional Properties of PFC, SMA and DGA Pavements. Proceedings of the th Transportation Research Board Annual Meeting (CD), Washington D.C., January 0. TRB 0 Annual Meeting CD-ROM