Asphalt Rubber Asphalt Concrete Friction Course Overlay as a Pavement Preservation Strategy

Transcription

1 Asphalt Rubber Asphalt Concrete Friction Course Overlay as a Pavement Preservation Strategy K. Kaloush, K. Biligiri, and M. Rodezno Department of Civil and Environmental Engineering, Arizona State University, Tempe, Arizona, USA M. Belshe FNF Construction, Tempe, Arizona, USA G. Way and D. Carlson Rubber Pavement Association, Tempe, Arizona, USA. J. Sousa Consulpav International, Inc. USA - Portugal ABSTRACT: The use of Asphalt Rubber (AR) pavements in the USA has been successful by several States. AR binder used in the hot mix asphalt is a mixture of 80 percent hot asphalt and 20 percent ground waste tire crumb rubber. Typically, AR-Asphalt Concrete Friction Course (AR-ACFC) mixes contain 9 to 10 percent asphalt rubber binder and their use has been primarily focused on reducing thermal and reflective cracking, and highway noise. This paper discusses the AR-ACFC benefits as a pavement preservation strategy. It highlights some results of the laboratory material characterization tests, and presents several field performance evaluation outcomes including: highway noise reduction, mitigation of daily thermal variances in Portland Cement Concrete pavements, improved skid resistance, reduced roughness, and reduction of emission rates of tire wear per kilometer driven. 1 ASPHALT RUBBER OVERLAYS IN ARIZONA The Arizona Department of Transportation (ADOT) began to use Open Graded Friction Courses (OGFC) with conventional asphalt as early as The primary reason for using this material was to provide a surface with good skid resistance and ride quality. During twenty years of Asphalt Rubber (AR) use, ADOT work evolved from using AR chip seals to utilizing reacted AR as a binder in open and gap graded asphalt concrete. In 1988, ADOT built its first AR mix project, which consisted of a 25 mm layer of an open-graded Asphalt Rubber Asphalt Concrete Friction Course (AR-ACFC) placed on Interstate 19 south of Tucson. This AR-ACFC mix was placed on top of a plain jointed concrete pavement and performed very well to date with very little maintenance. Hundreds of projects have been successfully built to date. A 50 mm structural gap graded Asphalt Rubber Asphalt Concrete (ARAC) overlay was designed and built in 1990 on Interstate 40 near Flagstaff, and it was overlaid with 12.5 mm AR-ACFC [Way 2000]. The purpose of the project was to compare the performance of the overlay on a severely cracked concrete pavement. The asphalt rubber sections built had the least percentage of reflective cracks, one third less than a 100 mm conventional overlay and less than one half a 200 mm overlay. Construction of an AR pavement involves first mixing and fully reacting the crumb rubber with the hot asphalt cement. Typically 20 percent ground tire rubber that meets specific gradation is added to hot asphalt heated to a temperature of about 190 C and mixed for at least one hour. After reaction the AR is kept at a temperature of about 175 C until it is introduced into the mixing plant. On cracked pavements, a gap-graded ARAC, generally 37.5 to 50 mm thick, is placed to address cracking. An AR-ACFC may be placed depending upon the traffic volume and type of highway. ARAC mixes are normally one half or less than one half the thickness of conventional Hot Mix Asphalt (HMA) pavements without asphalt rubber. The finished AR product is generally 25 to 50 percent more expensive, but is actually more cost effective when considered in a

2 life cycle cost analysis. This is because AR requires less maintenance costs (as shown in Figure 1a), has low percent cracking (Figure 1b), and therefore has longer service life. Maintenance Cost $/lane-km Overlays / Inlays 1000 AR-ACFC Year Percent Cracking HMA HMA AR AR Years Figure 1. Conventional Overlays Versus Asphalt Rubber: a) Maintenance Costs; b) Percent Cracking. This paper discusses the AR-ACFC benefits as a pavement preservation strategy. While there are several categories that distinguish the selection of a good pavement preservation alternative, the authors selected the following focus areas as the criteria for a good pavement preservation strategy: good mixture and binder characteristics through laboratory performance testing; reduced tire / pavement noise, improved thermal gradient characteristics and interaction with the urban climate; good field frictional characteristics for safety; improved ride quality and comfort for users; less tire wear emissions and therefore better impact on air quality; and finally, cost and energy demand effectiveness through life cycle cost analysis. 2 LABORATORY PERFORMANCE 2.1 Background The Department of Civil and Environmental Engineering at Arizona State University (ASU) has been involved with several asphalt rubber mixtures characterization studies. The studies are being conducted in cooperation with ADOT and other local and international transportation agencies. The ultimate goal from these studies is implementing a methodology for performance related specifications for asphalt rubber pavements, and developing typical design input parameters for specific conditions. In these characterization studies, hot AR mixes are obtained from the field during construction. The mixes are reheated and compacted at the air voids level specified for each project. The specific tests used for these studies are: consistency binder tests, repeated load (dynamic creep) for permanent deformation evaluation, Dynamic (complex) modulus for stiffness evaluation, flexural beam test for fatigue cracking evaluation and indirect tensile tests for thermal cracking evaluation (Kaloush et al, 2002, 2003a, 2003b, 2003c, 2004 ). For brevity, only the consistency binder and dynamic complex modulus tests are presented below for an indication of laboratory performance evaluation. 2.2 Binder Tests ADOT studies have shown that by using asphalt rubber as a binder, the film thickness is increased to a value of micrometer compared to the typical dense-graded HMA film thickness of about 9 micrometer [Way 2000]. In Arizona, the grade of asphalt binder used as a base to make AR is a PG58-22 (AC-10, Pen ), in contrast to the typically stiffer grade of PG (AC-20, Pen 60-70) used in the mountains. In the desert the AR base asphalt grade is PG (AC-20, Pen 60-70) compared to the PG (AC-40, Pen 40-50) typically used for dense graded mixes. The 20 percent ground tire crumb rubber particles change the AR tempera-

3 ture susceptibility as shown in Figure 2 [Kaloush 2002, Sotil 2003]. As it can be observed, the viscosity-temperature susceptibility of the rubber modified binders is better (flatter, lower slope) than the conventional (virgin) binder, both at high and low temperature conditions. At lower temperature conditions, the AR binders are softer than the virgin binder. Higher binder viscosities at high temperatures and lower viscosities at lower temperature are indicative of good overall mix performance characteristics. These characteristics also agree with observed field performance, where AR mixes are known to have better response against permanent deformation, and low-temperature cracking Viscosity (Log Log cp) ADOT Virgin PG I-17 AR PG I-17 AR PG I-40 AR PG Alberta AR Pen Temperature Rankine (R) Figure 2. Comparison of the Viscosity Temperature Relationships for the Different Binders. 2.3 Dynamic Complex Modulus E* Tests By current practice, dynamic complex modulus testing of asphalt materials is conducted per AASHTO TP (Witczak et al, 2002). Before presenting test results of the AR-ACFC mixes, a discussion on the state of stress applied in the laboratory should be presented. The E* tests can be performed either unconfined or with varying confinement levels. When comparing dense, gap and open graded mixes, confined dynamic modulus E* tests should be performed to appropriately rank the mixes. The confined Dynamic Modulus E* test is especially important for the open graded mixes because it represents the true state of stress of in the field (surface layer with high confinement stress under loading). The effect of confinement is clearly shown in Figure 3, where typical master curves for an AR-ACFC mixture test results are presented for unconfined and three levels of confinements: 69, 138, and 207 kpa. The confined test results yield much higher moduli and the difference among the level of confinements continue at high temperatures. Table 1 shows the complex modulus test results for several AR mixtures compared to conventional mixtures (by means of a modular ratio comparison). The results show that the AR-ACFC mixes have better moduli values compared to the conventional mixes. A comprehensive documentation of the test results on the various asphalt rubber mixtures tested can be found in several publications by the authors [Kaloush 2002, 2003a,b,c and Sotil 2003].

4 100,000 AR-ACFC-Unconfined Confined 69 Kpa Confined-138 Kpa Confined-207 Kpa 10,000 E* (Mpa) 54.4 C 0.1 Hz 1, Log Reduced Time (sec) Figure 3. Comparison of E* Master Curves and Effect of Confinement for Typical AR Mixture. Table 1. Typical Modular 37.8 o C / 10 Hz for AR and Conventional Mixes, Confined Testing. Mix ID Binder Type AC % Va % Nom. Aggregate Modular Ratio ARAC (stiff) (AR) mm GG 1.08 AR-ACFC (AR) mm OG 1.02 ARAC (AR) mm GG 1.00 Conventional mm DGM 0.94 Conventional AC mm DGM 0.77 ARAC (Soft) Pen (AR) mm GG 0.67 OG = Open Graded Mixture DGM = Dense Graded Mixture GG = Gap Graded Mixture 1 Reference Mix 3 TIRE / PAVEMENT NOISE CHARACTERISTICS Highway traffic noise is generated by engine, exhaust, and tire / pavement interaction. It has been well documented in the literature that the tire / pavement interaction is the dominant source of highway noise. Dominant factors that contribute to the tire / pavement noise include: air pumping, compression of tread block, friction, porosity, absorption, aggregate texture, thickness, age of the pavement and temperature (Biligiri et al, 2008). There are different views among experts on which are the dominant factors or mechanism causing tire/pavement noise. The authors believe that AR-ACFC mixes reduce tire/pavement noise because they act as an acoustic absorber due to the viscoelastic nature of the asphalt mix, and because air pushed through the layer voids (> 18%) avoiding the air compression under the tire. Furthermore, the smooth riding surface characteristics and small top size aggregate contribute to less tire deformation with travel and less squeezing of air between the tire and pavement. The viscoelastic characteristics of an AR-ACFC come from much higher asphalt binder content (9-10%) and inclusions of crumb rubber (20% by weight of the binder). When comparing field noise measurements of the rubber versus non-rubber ACFC mixes (as shown in Figure 5), there is reason to believe that rubber particles in the AR-ACFC contribute to less noise due to the sound absorptive characteristic of rubber materials.

5 Asphalt rubber friction course mixes were shown to be more noise dampening than conventional and other modified mixes in terms of lower field noise measurements. ADOT conducted a pavement preservation experiment on the Interstate 10 (I-10) in 1999 (Scofield 2000). As part of this experiment, 32 replicate test sections were constructed constituting five asphalt concrete pavement wearing courses. The Annual Daily Traffic (ADT) for this highway is about 60,000 with 25% trucks. The five different pavement types included: Permeable European Mixture (PEM), Stone Matrix Asphalt (SMA), Asphalt Rubber Open Graded Friction Course (AR- ACFC), Polymer Modified Open Graded Friction Course (P-ACFC), and ADOT s Standard Open Graded Friction Course (ACFC). The AR-ACFC mix experienced the least cracking and wear after eight years of service with the other test sections showing considerable cracking and wear as illustrated in Figure 4. Figure 4. Example of Pavement Surface Condition in 2007 for Two Pavement Types after Eight Years of Service. On-Board Sound Intensity (OBSI) noise measurements were taken during the Fall 2002 by ADOT as part of the Arizona s Quiet Pavement Program. In addition, Dynatest Inc. obtained new noise measurements in March 2008 as part of a larger California Arizona highway noise study [Scofield 2000, 2003, CALTRANS 2006]. Figure 5 shows a comparison of average noise readings between the two measurement periods for the five pavement types. The least noise observed for both periods is for the AR-ACFC mixture. This difference agrees with visual distress observations made in the Fall of Several sections that exhibited higher noise have greater amount of raveling and cracking. Tire / Pavement Noise (db) AR-ACFC ACFC P-ACFC PEM SMA Dynatest 2008 at 100 Km/h Scofield-Donovan 2002 at 100 Km/h Figure 5. Comparison of Average Tire / Pavement Noise (db) for Arizona I-10 Test Sections.

6 4 TEMPERATURE GRADIENT EFFECT ON PORTLAND CEMENT CONCRETE Since 2003, ADOT has been placing AR-ACFC mixes over existing Portland Cement Concrete Pavements (PCCP) as part of the Quiet Pavement Program. As of January 2008, approximately 200 kilometers of urban freeway kilometers have been overlaid in the greater Phoenix area. Recently, joint ASU-ADOT studies have been conducted to evaluate the consequences of this paving strategy on the Urban Heat Island (UHI) effect, and the insulating effects of the AR-ACFC mix on PCCP. Temperature data collected over the past 4 years showed that the darker AR-ACFC surface color increases the surface temperatures of the pavement during daytime. However, the nighttime UHI effect showed a benefit of using the AR-ACFC overlay in reducing the pavement surface temperatures due to the porosity and lower thermal mass of the layer [Golden and Kaloush 2006]. An important consideration is subjecting these surfaces to traffic, which provides the necessary aeration effect. These findings are discussed below. Several studies have shown that temperature-induced pavement responses are more significant than traffic induced responses (Mahboub et al, 2004). Rigid pavements are affected by a range of environmental factors including temperature. The daily temperature variations generate curling stresses which can ultimately lead to a partial loss of subgrade contact. Temperature gradients in particular in a rigid pavement slab have been shown by some researchers to directly relate to fatigue damage (Masad et al, 1996). A gradient of 0.56 C/25 mm in the slab increases fatigue damage due to truck traffic by a factor of 10 as opposed to a zero-temperature-gradient condition. Interstate 10 at Ray Road in the Phoenix area offered an opportunity to create excellent modeling conditions for quantifying the effects of AR-ACFC overlays of PCCP. The I-10 pavement north of Ray had been overlaid with AR-ACFC as part of the Quiet Pavements Phase 3 project in May With two test sites within the vicinity of the start of the overlay, one within the AR-ACFC overlay area and one with only PCCP, pavement temperature data were generated for both conditions of PCCP with and without AR-ACFC (see Figure 6). Also, each site included temperature sensor placement located in the shoulder areas of each condition allowing a matrix that included data from areas with and without traffic. In May 2006 the Ray Road I-10 test site was constructed. The existing PCCP was cored and sensors were placed using dowels to insure top to bottom spacing. Sawcut lines were then made from the cores to the shoulder so that future data collection could be made without impacting traffic. Data was recovered for the month of June 2006, and the data were examined in depth. Figure 6. Ray Road Test Site at I-10.

7 Table 2 shows comparisons of the average temperature differentials (ΔT) measured for June Curling stresses for each respective section were calculated. The section with traffic and without the AR-ACFC overlay experiences daytime induced stresses on the magnitude of 25% greater than the section with traffic and with the AR-ACFC overlay. Night time values for the section without AR-ACFC were about 8% higher. When considering that a major portion of the damage to a PCCP structure is thought to result from thermal gradient induced stresses as opposed to traffic loadings, the service life of the PCCP can be significantly extended with the use of AR-ACFC overlays as a pavement preservation strategy. Table 2. Comparison of Temperature Differentials. Overlay / Traffic Case Max ΔT C Min ΔT C Range of ΔT C With AR-ACFC: Traffic vs. No Traffic Without AR-ACFC: Traffic vs. No Traffic With Traffic: AR-ACFC vs. No AR-ACFC Without Traffic: AR-ACFC vs. No AR-ACFC ROUGHNESS Pavement roughness is measured by ADOT Pavement Management Division using a profilometer. The results are reported in International Roughness Index (IRI). The IRI provides a numeric scale of measuring roughness with a break point between what is considered rough and smooth pavement is often considered to be about 2 m/km. IRI values less than 1 m/km are considered very good by Federal Highway Administration guidelines. Typical IRI measurements before and after the AR-ACFC overlay on PCCP pavement sections in the Phoenix area are shown in Figure 7. It is observed that the AR-ACFC overlays provide a substantial improvement in ride quality by reducing roughness in half (on the average). Average IRI (m/km) PCCP (Before) AR-ACFC (After) Pavement Section Figure 7. Comparison of Roughness Measurements Before and After the AR-ACFC Overlay. 6 FRICTION Surface friction is also measured by ADOT Pavement Management Division utilizing a MU Meter. Measurements are reported as a skid number (measured value of friction times 100). In Arizona, the intervention level for friction reported for interstate, primary and secondary roadways is Similar to the roughness measurements, typical average friction values before and after the AR-ACFC overlay on PCCP pavement sections in the Phoenix area are shown in Figure 8. As expected, and because this is a wet friction test, the data shows an improvement in the skid numbers after the AR-ACFC overlay (except for pavement section 7). In addition, the fric-

8 tion data after the overlay are more uniform as the overlay seems to correct polished surface problem that existed on the original PCCP pavement. Average Friction Value (mu) PCCP (Before) AR-ACFC (After) Pavement Section Figure 8. Comparison of Friction Measurements Before and After the AR-ACFC Overlay. 7 TIRE WEAR EMISSIONS Tire wear contributes to atmospheric Particulate Matter (PM) and is regulated by the United States Environmental Protection Agency because PM has been shown to affect human health. Vehicle emissions are a significant source of both PM2.5 and PM10. Vehicle fleet emissions per mile traveled have been reduced significantly in the last 30 years as a result of improved engine operation and tailpipe controls. However, zero emission vehicles will continue to generate PM from tire wear, road wear, brake wear, and re-suspended road dust. In a 2005 ASU study, aerosol measurement techniques were applied to evaluate tire wear emissions from the vehicle fleet using the Deck Park highway tunnel in Phoenix, Arizona. The Deck Park Tunnel highway surface was PCCP, and was resurfaced with an AR-ACFC layer as part of the ADOT Quiet Pavements Program. This study took advantage of a rare opportunity to sample tire wear emissions at the tunnel before and after the AR-ACFC overlay. The study reported on the measured PM emissions from the on-road vehicle traffic during typical highway driving conditions for the two different roadway surfaces. It presented the analysis of representative tire tread samples for tires wear marker compounds. The tire wear analysis of representative tire tread samples included tire wear marker compounds, which have been identified and quantified in representative tire wear composite samples from used tires in Arizona, ambient aerosol samples and aerosol samples collected from the two roadway tunnels. Test extraction and separation protocols to determine the amounts of tire wear tracers in tire treads have been also developed as part of this study [Alexandrova et al, 2007]. The emission rates of tire wear tracer compounds (identified as #3 and #4) have been calculated as shown in Table 3. Emission rates of tire wear tracers were found higher at the PCCP road surface than at AR-ACFC road surface. The emission rates of tire wear per kilometer driven at PCCP road surface were times higher than emission rates of tire wear at AR- ACFC road surface. These findings provided ADOT with revised tire wear emission data for use in their federally-mandated air quality modeling for the Phoenix airshed. Table 3. Tire Wear Emission Rates Measured in the Deck Park Tunnel, μg/km.

9 Tire wear emission Experiment 1 Experiment 2 rate based on (PCCP road surface) (AR-ACFC road surface) Compound # ± ± 35 Compound # ± ± 24 Compounds 3 and 4 are tire wear tracers. 8 COST AND ENERGY CONSIDERATIONS As it was mentioned earlier, the finished AR product is generally 25 to 50 percent more expensive. However, life cycle cost analysis has shown that a substantial dollar savings can be obtained over the expected life of a project when AR paving strategies are employed [Hicks and Epps, 2000]. This is because AR requires less maintenance costs, has low percent cracking, and therefore has longer service life. The energy savings by using AR (both gap graded and open graded designs) is quite impressive [Sousa et al, 2006]. Table 4 represents the heat of combustion values for crumb rubber modifier used in AR. The kj/kg represents energy savings in terms of less asphalt concrete overlays is needed over the life of the pavement. It is also the result of using less than one half the thickness of conventional paving material with equal or better field performance. The kj/kg energy savings refers to a two inch AR gap graded overlay used as an alternative to a conventional four inch asphalt pavement overlay. This has been a common practice of ADOT for over 135 pavement preservation projects built to date. The kj/kg energy savings refers to one inch open graded AR-ACFC mix used on top of a Portland cement concrete pavement in comparison to a five inch conventional asphalt overlay. The kj/kg energy savings refers to the mining energy and transport energy associated with using thicker pavements compared to the thinner AR pavements. Table 4. Energy Utilization (kj/kg) for Asphalt Rubber [Sousa et al, 2006]. Process Energy Gain / Loss (kj/kg) Tire Shedding Shred Transportation Granulation CRM Transportation Steel Recovery Asphalt Saved +209,325 to 465,168 Aggregate Saved +107,860 Total Gain / Loss +310,267 to 566,109 9 CONCLUSIONS This paper discussed specific aspects on why AR-ACFC overlays are considered as good pavement preservation strategy. The focus areas summarized the outcome of several research studies. Many benefits have been identified for the utilization of AR-ACFC overlays. The reduction of cracking has been confirmed through field observations and laboratory testing conducted on the binder and the mixture. The restoration of a smoother ride and the increase in skid resistance were additional benefits identified through several pavement test sections. There is no doubt that AR-ACFC overlays provide the least tire/pavement noise compared to any other road surface type evaluated in these studies. In addition, AR-ACFC overlays have shown to have a significant impact on reducing the induced stresses in PCCP due to thermal gradients. A composite PCCP/AR-ACFC pavement design will be very long lasting. Emission rates of tire wear per kilometer driven on AR-ACFC overlays were reduced by half, having great impact on air quality and human health in urban areas. Life cycle cost analysis and energy considerations were also favorable for this pavement preservation strategy.

10 10 REFERENCES Alexandrova, O., Kaloush, K. and Allen, J Impact of Asphalt Rubber Friction Course Overlays on Tire Wear Emissions and Air Quality Models for Phoenix, Arizona Airshed. Journal of the Transportation Research Board, No. 2011, pp Washington, D.C. Biligiri, K. P., Kaloush, K. E., and Golden, J. S Paving Material Properties and Tire / Pavement Noise, accepted for Publication in Eurasphalt & Eurobitume (E&E) Congress 2008, Copenhagen, Denmark. Golden, J. S. and Kaloush, K.E. 2006, Mesoscale and Microscale Evaluation of Surface Pavement Impacts on the Urban Heat Island Effects. International Journal of Pavement Engineering, Volume 7, No. 1, pp 37-52, Taylor & Francis. Hicks, R.G. and Epps, J. A Life Cycle Costs For Asphalt Rubber Paving Materials, Proceedings of the Asphalt Rubber 2000 Conference, Vilamoura, Portugal. Kaloush, K. E., Witczak, M. W., Way, G. B., Zborowski, A., Abojaradeh, M., and Sotil, A Performance Evaluation Of Arizona Asphalt Rubber Mixtures Using Advanced Dynamic Material Characterization Tests. Final Report, Arizona State University, Tempe, Arizona. Kaloush, K. E., Zborowski, A., and Sotil, A. 2003a. Performance Evaluation of Asphalt Rubber Mixtures in Alberta. Final Report, Arizona State University, Tempe, Arizona. Kaloush, K., et al. 2003b. Material Characteristics of Asphalt Rubber Mixtures. Proceedings Asphalt Rubber 2003 Conference, ISBN , p , Brasilia, Brazil. Kaloush, K.E., Witczak, M. W., Sotil, A., and Way, G. B. 2003c. Laboratory Evaluation of Asphalt Rubber Mixtures Using the Dynamic Modulus (E*) Test. CD-ROM Publication Annual Transportation Research Board Meeting. Washington, D.C. Kaloush, K., et al., Cracking Characteristic of Asphalt Rubber Mixtures, Proceedings of the Fifth International Rilem Conference, PRO 37, Pages , Limoges, France, Mahboub, K. C., Liu, Y., Allen, D., Evaluation of Temperature Responses in Concrete Pavement, Journal of Transportation Engineering. Masad, E., Taha, R., Muhunthan, B., Finite element analysis of temperature effects on plain-jointed concrete pavements, Journal of Transportation Engineering. Scofield, Larry, Arizona s Pavement Preservation Experiment, Arizona Department of Transportation, Phoenix, Arizona. Scofield, L. and P. Donovan, Development of Arizona s Quiet Pavement Research Program, Proceedings of Asphalt Rubber 2003, Brasilia, Brazil. Sotil, A Material Characterization of Asphalt Rubber Mixtures Using the Dynamic Modulus Test. Master of Science Thesis. Arizona State University, Tempe, Arizona. Sousa, J., Way, G. B. and Carlson. D. D Energy Consumption of Alternative Scrap Tire Uses, Proceedings of the Asphalt Rubber 2006 Conference, Palm Springs, California. The California Department of Transportation, Division of Environmental Analysis, Further Development of the Sound Intensity Method of Measuring Tire Noise Performance of In-Situ Pavements, page 9, January 2006.

11 Way, G. B., Flagstaff I-40 Asphalt Rubber Overlay Project nine years of success. TRB 79th annual meeting, Washington D.C. Way, G. B., 2000a Asphalt Rubber-Research and Development. 35 years of progress and controversy, ETRA annual meeting, Brussels. Belgium. Witczak, M. W., Kaloush, K. E., Pellinen, T., El-Basyouny, M., & Von Quintus, H Simple Performance Test for Superpave Mix Design. NCHRP Report 465. Transportation Research Board. National Research Council. Washington D.C