Porous Elastic Road Surface as an Urban Highway Noise Measure

Transcription

1 Porous Elastic Road Surface as an Urban Highway Noise Measure Seishi Meiarashi, Public Works Research Institute, 1-6 Minamihara Tsukuba Ibaraki Japan Abstract. Highway traffic noise in urban areas of Japan is a serious problem, not only for residents along highways, but also for highway administrators. Only 13 percent of urban highways have met the environmental standard for noise. Noise barriers cannot be used as a noise countermeasure on the majority of highways on which access is not controlled. This problem is impeding new highway construction in urban areas. The Public Works Research Institute (PWRI) has, since 1993, been developing a new low-noise pavement named Porous Elastic Road Surface (PERS). This new pavement has a porous structure composed of granulate rubber made from old used tires as its aggregate, and urethane resin as its binder. The author estimates that the potential noise reduction levels in Leq exceed 10 db(a). More than 90 percent of highways in urban areas would meet the standard if this noise reduction level were achieved. This paper examines the general performance of PERS obtained through past development at the PWRI. It also summarizes the results of recent research done to further improve the noise reduction levels of PERS, and the first test construction using PERS in Japan. As a result, the noise measurement result at the site indicates that PERS has an enough potential to solve highway traffic noise problems in urban areas of Japan. INTRODUCTION The Public Works Research Institute (PWRI) has, since 1993, been developing a new low-noise pavement named Porous Elastic Road Surface (PERS). This new pavement has a porous structure composed of granulate rubber made from old used tires as its aggregate, and urethane resin as its binder. Its porosity is approximately 40 percent. The pavement was first proposed in Sweden in the 1970 s, however, Swedish researchers have failed to improve it as a practical pavement. Noise reduction levels are 15 db(a) for cars and 8 db(a) for trucks. The author estimates that the potential noise reduction levels in Leq exceed 10 db(a). More than 90 percent of highways in urban areas would meet the environmental standard for noise shown in Table 1, if this noise Table 1 Environmental Standard for Arterial Highway Noise Day Time Night Time 70 db 65 db reduction level were achieved. The PWRI has already solved several of the problems with PERS, for example, its insufficient adhesion between the pavement and the base course, its low skid resistance, and its poor fireproof performance. Its technical level has already reached the stage of test construction on urban highways. This paper examines the general performance of PERS obtained through past development at the PWRI. It also summarizes the results of noise reduction levels of PERS at the first test construction site in Japan. The first part deals mainly with the improvement of the noise reduction effect by changing its porosity and thickness, adhesion to the base course, durability, wear resistance, wet friction, and fire resistance ; whereas the second part focuses on the laboratory performance testing in advance, to identify a new construction method of PERS before trial construction on highways, and the noise reduction effect observed at the construction site. LATEST TECHNOLOGY Noise reduction Noise reduction is the most interesting aspect of PERS, and the author has examined this feature in each of the four specifications of PERS that have been released. In the first test construction, its porosity was 40% and its thickness was 5 cm. Power levels of vehicles were measured by the controlled pass-by method based on ISO 326 and ISO The detailed methodology to calculate power levels was described by Meiarashi (1). Figure 1 illustrates that for all vehicles, PERS is superior to Drainage Asphalt Pavement (DAP). The superiority is expressed by differences of the total A-weighted sound power levels for constant speed. As compared to porous 1

2 Figure 1 Power levels measured at PWRI test course in 1994 : 5 cm thick, 35% of porosity, L=25m, W=5m asphalt pavement (Dense Asphalt Pavement (DENAP)), the noise reduction attained with PERS is from 2 to 10 times greater than that attained with DAP. Note that the noise reduction for a car is 13 db(a) at 60 km/h, and 6 db(a) for light and heavy trucks. For cars, the coast-by noise is the dominant contribution to power-by noise, whereas for trucks it is the power-unit noise. Thus, there is a clear difference in noise reduction between cars and trucks, owing to the relatively large power-unit noise of trucks. The axial weights on the light truck and heavy truck were 10 tons and 5 tons, respectively. The author has conducted four noise measurements in total at the PWRI testing course to improve the noise reduction effect of PERS, including the first one described above. The second noise measurement in 1995 was focused on the influence of porosity on noise reduction. Figure 2 shows that noise reduction of PERS is almost (a) Heavy Truck (c) Passenger Car Figure 2 Power levels and PERS porosity (b) Light Truck 2

3 saturated at the porosity of 35% and over. In the third noise measurement of 1996, a major issue was the effect of PERS thickness on noise reduction. The optimal PERS noise reduction levels for passenger cars, light trucks, and heavy trucks are db(a), 4-5 db(a), and 3-5 db(a), respectively. Figure 3 reveals that the noise (a) Heavy Truck (b) Light Truck (c) Passenger Car Figure 3 Power levels and PERS thickness reduction of PERS is at a maximum at the thickness of 3 cm. Considering the relatively small difference of noise reduction between 3 cm thickness of PERS and 2 cm thickness of PERS, and material cost reduction, the optimal thickness of PERS seems to exist between 2 cm and 3 cm. The optimal PERS noise reduction levels for passenger cars, light trucks, and heavy trucks are db(a), 8-9 db(a), and 6-10 db(a), respectively. The fourth noise measurement was conducted to confirm the noise reduction effect of PERS, which met the final criteria for PERS, obtained in the cooperative research of PWRI and private rubber product companies. One of the serious issues to solve in this research was to improve the low wet friction of PERS. Almost all the companies had changed the component of PERS. The noise reduction levels of PERS on the passenger car are db (A) excluding only one product whose reduction level is 8-9 db (A). Those of PERS on trucks are 8-10 db (A). As a result, the change in the component specification by the companies had to improve wet friction while sacrificing noise reduction for passenger cars. Adhesion to base course Polyurethane adhesive between PERS and semi-flexible pavement as base course showed insufficient performance. In 1994, passage by a heavy truck caused PERS to peel off from the base course. In 1997, the 3

4 author identified the adhesion performance criterion as 0.8 MPa for PERS, after moisture and heat accelerated deterioration tests through both analytical and numerical calculations, and found that epoxy resin adhesive, which showed much stronger adhesion than polyurethane resin in the two-face shear test, satisfied this criterion. Durability Accelerated pavement tests, as illustrated in Figure 4, were conducted from 1994 to The total cumulative traffic volume of test cars finally reached 180,000, corresponding to a 1.2-month exposure to an ordinary highway whose heavy traffic volume is 3,000 per day, per lane. Figure 5 shows the result of maximum rutting depth. PERS shows better performance than DAP, with far better deformation performance than conventional pavement such as DENAP. Figure 4 Accelerated pavement test Figure 5 Rutting depth after accelerated pavement test Wet friction Low wet friction had been a serious issue of PERS from the initial development stage. The component of PERS was changed through cooperative research between PWRI and rubber product companies to solve this problem, as mentioned in the previous section. Different kinds of additives, for example, broken pieces of pet bottles or small particles of specific inorganic matter, improved the skid resistance. Figure 6 shows the results of the wet friction measurement test described in Figure 7. The car shown in Figure 7 provides a special tire for measuring skid resistance. A load-cell detects the skid resistance force caused between the tire and pavement surface. The signal generated at the cell is amplified at a strain meter, and then the amplified signal is recorded by an analogue type of pen recorder. The black dotted line, R.S.O., means the minimum criterion of wet friction regulated in the technical guideline for highway design in Japan. 4

5 100 Figure 6 Wet skid resistance Figure 7 Apparatus for wet skid resistance Fire resistance Fire resistance was thought to be a potential problem, since rubber may burn fiercely. The fire hazard problem has been studied by PWRI. Squares of PERS 5 5 m were placed outside a laboratory, 36 liters of diesel oil or gasoline were sprinkled on the surface, as well as on an adjacent (conventional) asphalt pavement. The fluid was then ignited with a torch, and factors such as pavement materials, height of flames and generation of smoke were observed and the tests were also filmed. In the experiments, three surfaces were compared: dense asphalt concrete, porous asphalt concrete and the 5 5 m panels of PERS. The results, as given in Table 2(2), show that regarding spreading speed and flame height, the PERS was safer than the dense asphalt concrete. Figure 8 illustrates these tests. 5

6 Table 2 Fire resistance test condition Surface type DENAP DAP PERS Burning of fuel and pavement materials Fuel oil spreading over the pavement surface strongly burned with reddish flames but the pavement did not burn. Fuel oil evaporated through the voids of the pavement ignited, causing blue flames. However, pavement materials did not burn. Fuel oil evaporated through pavement voids ignited; rubber panels burned up, causing reddish flames. Fire spread over the pavement very slowly. Flame height Smoke generation m Fuel oil burned incompletely, producing a column of black smoke. Approximately 0.3 m Only a little smoke was observed m A column of black smoke was observed from the burning rubber panels. (a) Dense Asphalt Pavement (b) PERS Figure 8 Fire resistance 6

7 FIRST TEST CONSTRUCTION Pavement Structure & Construction PERS construction in highways requires the structure to be developed as a total pavement system and a construction method that is very different from the previous ones in PWRI test courses. There are two reasons for improving the structure and construction method. The first one is a time constraint. One potential application of PERS is for heavy-traffic arterial highways in urban areas, where the working time is limited to 10 hours at night (such as from 8 PM to 6 AM) to avoid causing traffic congestion. The standard area of pavement resurfacing of an urban highway is 2,000-3,000 square meters per day. The construction work involves removing the existing wearing course & base course, constructing new semi-flexible pavement as the base course, putting adhesive on the base course, and paving PERS as shown in Figure 9(a). Considering the working time before paving PERS, it is impossible to complete all the works within the time limit. (a)conventional PERS,3cm (b)test Construction PERS,3cm Semi-flexible Pavement, 5cm ILB, 8cm Base Course Hard Type of Epoxy Resin Adhesion (1kg/m 2, On-site Execution) Asphalt Coated Sand, 2cm Base Course Geotextile 60g/m 2 Figure 9 Pavement Structure The second reason is for quality control of adhesive performance. In the early stage of development of PERS, there were various troubles concerning the adhesive as mentioned in the previous section. The polymer type of adhesive is very sensitive to the ambient conditions of curing such as temperature and humidity. It seems very difficult to maintain stable performance of the adhesive during outdoor work. In response to these problems, pre-fabricated types of PERS would appear to be the only solution. The main prefabricated types of pavement products are Inter-Locking concrete Block (ILB), Pre-stressed Concrete Panel (PCP), and Reinforced Concrete Panel (RCP). The ILB has been widely used for pedestrian ways especially in prestige areas and shopping malls, where architects and planners are interested in the visual impact of paving. Some ILB s are also to be found in industrial areas, such as storage yards and dock-side paving, where the main concerns are structural performance, cost and maintenance. The PCP is pre-tensioned in the transverse direction during fabrication, and post-tensioned together in the longitudinal direction after placement. The PCP and RCP are mainly used for sections where extremely high durability is required, such as the pavement in tunnels. (3),(4) In view of the time constraints, it is impossible to use PCP and RCP as the base course of PERS due to the slow speed of construction of less than 100 square meters per ten hours. The present mechanical method of laying ILB improves the construction efficiency and overcomes the constraint. With this background, the author has proposed using ILB for the first test construction of PERS. However, ILB has been used in very few cases for highways and its durability for the surface course is unknown. The author has clarified the initial durability of the ILB-PERS composite surface by accelerated pavement tests in the laboratory shown in Figure 9(b). No fatal damage to the surface was found after 12,000 passes of the test truck. 7

8 PERS was first constructed in Tazawa on National Highway Route 46, on the 18 th of October, The total number of lanes is two, and the width of each is 3.75 meters. The total length is 20 meters. The traffic volume, heavy traffic ratio, and speed limit are; 120 vehicles per day, 20%, and 60km/hr, respectively. Figure 10 shows the general view of the section and the initial condition of the PERS surface. Figure 10 First test construction site Table 3 shows the requirement time for this test construction. A former work time to prepare the base course of PERS was five hours, and the settlement of ILB by human handwork was six hours. The total construction time was eleven hours. Construction work by an exclusive construction machine will save the latter half of working-time, however, it would be very difficult to save the former half of the time for removing the existing pavement surface. The total depth of replacing the existing surface is 28 cm, and it is impossible to temporarily open to public traffic. It means that the former and latter work should be done continuously. Table 3 Test construction work Classification Specification Working-Time (hr) Traffic Regulation 1 Surface Milling Surface Depth: 18cm 0.75 Base Course Depth: 10cm Total Area: 75m2 Base Course Treatment 0.5 Side-Edge Concrete Block 1.25 Upper Base Course Base Course Thickness: 15cm 1 Drain System Total Length: 28m 0.5 Subtotal 5 ILB Total Area: 75m2 6 Total 11 The test construction results strongly indicate that the thinner and higher bending strength of the concrete base for PERS, should be able to reduce the construction work time. Noise Measurement The Finite Sound Integrated Method proposed by Meiarashi (1) was applied for noise measurement to calculate power levels of individual vehicles. The vehicles were limited to smaller ones such as passenger cars and light trucks, because of the short section length. Figure 11 illustrates the arrangement of equipment, including a sound level meter as a microphone and two sets of photo-detectors as a speed meter. Figure 12 shows the A-weighted 8

9 ower levels of vehicles measured on PERS and DENAP. When noise reduction levels are defined as the difference in the levels between PERS and DENAP, they are approximately given by the following formula: PWL = log 10 V PWL: Noise reduction level db (A) V: Vehicle speed (km/hr) Figure 11 Equipment for noise measurement Figure 12 Power levels measured at first test 9

10 AP 100,000 Figure 13 Power spectrums These noise reduction levels seem to be smaller than those measured at the PWRI test course. The difference in geometric relationship between vehicles as noise sources and the microphone may cause this discrepancy. The distance between vehicle running position and a microphone in the highway area was shorter than that of noise measurement in the PWRI test course, which means that additional attenuation loss could not be expected in the field measurement. The lower reflecting angle of the site than that of the PWRI test course might be another factor for the lower noise reduction of PERS. In general, lower sound reflecting angles provide smaller sound absorption coefficients. Figure13 shows the power spectrum calculated by measured noise data by using the 10

11 regression analysis of the relation between the vehicle speed and all the 1/3-octave bandwidth of the power spectrum, in order to adjust the difference in the vehicle speed measured on DENAP and PERS. The vehicle speed of this power spectrum is 80(km/hr). All-path power level indicated by AP in the figure means the sum mention of the entire 1/3 octave power spectrum. The noise reduction effect of PERS was obviously found in the 1/3-bandwidth frequency of 800Hz and over. Wet friction Figure14 shows the skid resistance measured at the site. This resistance was also measured on the same equipment mentioned in the previous section related to the skid resistance. The resistance of PERS becomes smaller than those of dense asphalt pavement in all vehicle speeds. The difference in skid resistance between PERS and dense asphalt pavement becomes smaller, as the vehicle speed becomes faster. Figure 14 Wet skid resistance of ILB type PERS Repair and withdrawing Pavement surface condition had been traced for three months after PERS was open to the public. Figure15 shows the maximum vertical displacement between ILB. PERS caused no displacement after just ten days, however, the maximum displacement became 5mm after 25 days. This displacement was mainly concentrated in the wheel paths of heavy vehicles. At this period, all the ILB at this portion were removed and loss in the cushion sand was detected. This tracing had been continuing after the repair of PERS by supplying the cushion sand, while the displacement had been increasing. The displacement exceeded the allowable limit and the dense asphalt pavement replaced all the ILB s only on the wheel path portion, which caused the displacement after 110 days. Figure 16 illustrates the damaged pavement surface just before the replacement. Finally, the highway administration office withdrew all the ILB type of PERS approximately after one year. Consideration on the factor, of what caused this displacement, finally came to the conclusion that the major reason was the insufficient drainage capacity of this section. When the excessive amount water, which is more than the drainage capacity, comes into the cushion sand layer under ILB, all the pores of the layer are filled with water. When the vehicle wheel passes on ILB, the water pressure increases and this pressure pushes the sand out 11

12 of the layer, through the space between ILB. The displacement suddenly happened just after snow started falling. The snow, which fell and accumulated on the road pavement during the night, rapidly melted in the morning as the air temperature had been increasing. This phenomenon caused the large amount of water, which saturated the void of cushion sand layer. The drainage capacity of ILB was designed and constructed by the technical guidelines of inter-locking concrete pavement. This result suggests that the guideline probably does not reflect the specific climate condition in cold snowy areas, where they have a lot of snow during the winter season. Figure 15 Vertical displacement of ILB Figure 16 ILB damage 12

13 It also means that this damage is an essential problem of ILB pavement, but not of PERS. Inefficient construction work and insufficient durability in the ILB type of PERS, suggests the necessity of developing a new base course block structure of PERS. CONCLUSIONS The Public Works Research Institute (PWRI) has, since 1993, been developing a new low-noise pavement named Porous Elastic Road Surface (PERS). The author estimates that the potential noise reduction levels in Leq exceed 10 db(a). The PWRI has already solved several of the problems with PERS such as insufficient adhesion between the pavement and the base course, low skid resistance, and poor fireproof performance. Based on the above mentioned research results, PERS was first constructed on the National Highway Route 46. Noise reduction levels measured in the field were less than expected, because the size of the construction area was very small. The noise reduction levels measured at the site seem to be a little bit smaller than those measured at the PWRI test course. The noise reduction effect of PERS was obviously found in the 1/3- bandwidth frequency of 800Hz and over. The resistance of PERS becomes smaller than those of dense asphalt pavement in all vehicle speeds. The difference in skid resistance between PERS and dense asphalt pavement becomes smaller as the vehicle speed increases. The highway administration office withdrew all the ILB type of PERS after approximately one year. Inefficient construction work and insufficient durability in the ILB type of PERS suggest the necessity of developing a new base course block structure of PERS. REFERENCES (1)Meiarashi S, et al. (1996): Noise Reduction Characteristics of Porous Elastic Road Surface, Applied Acoustics, Vol. 47, No. 3, pp (2)Ulf Sandberg and Jerzy A. Ejsmont (2002): Tire/Road Noise Reference Book, INFORMEX Ejsmont & Sandberg Handelsbolag (3)Alan Lilley (1988): Precast Concrete Paving History, Design, Applications and Problems, The Journal of the Institute of Highways and Transportation, pp (4)David Merritt, B. Frank McCullough, and Ned H. Burns (2001): Feasibility of Using Precast Concrete Panels to Expedite Construction of Portland Cement Concrete Pavements, Transportation Research Record 1761, Paper No