Field Evaluation of Noise Reducing Pavement: A Controlled Experiment. prepared by

Field Evaluation of Noise Reducing Pavement: A Controlled Experiment prepared by Yuen-Ting Fiona Leung, M.A.Sc. Candidate Corresponding Author University of Waterloo, 200 University Ave. West, Waterloo, ON., Canada, N2L 3G1 (519) 888-4567 ext. 3872 (Office) yfleung@engmail.uwaterloo.ca Susan Tighe, Ph.D., P.Eng. Canada Research Chair in Pavement and Infrastructure Management, Associate Professor of Civil Engineering University of Waterloo, 200 University Ave. West, Waterloo, ON., Canada, N2L 3G1 (519) 888-4567 ext. 3152 (Office), (519) 888-4300 (Fax) sltighe@civmail.uwaterloo.ca Scott Penton, P.Eng. Noise Specialist Associate RWDI Air Inc. 650 Woodlawn Road West, Guelph, ON., Canada, N1K 1B8 (519) 823-1311 (Office), (519) 823-1316 (Fax) scott.penton@rwdi.com Gary MacDonald, P.Eng. Head of Transportation, Design and Construction Division Regional Municipality of Waterloo 150 Frederick Street, Kitchener, ON., Canada, N2G 4J3 (519) 575-4721 (Office), (519) 575-4430 (Fax) mgary@region.waterloo.on.ca Paper Offered for Presentation and Publication To AISIM Waterloo, August 6, 2005 Submission Date: July 8, 2005 Word Count: 4371 words + 12 figurestables = 7371 words

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 2 ABSTRACT Traffic noise is a growing concern as traffic demand increases with urban sprawl. The most common traffic noise mitigation methods include the construction of noise barriers and earth berms. Although effective, these methods can be expensive and in the case of earth berms, they require large areas. The Regional Municipality of Waterloo in Ontario, Canada, in partnership with the University of Waterloo Centre for Pavement and Transportation Technology, is investigating the feasibility of reducing traffic noise by using different kinds of asphalt pavements, namely Stone Mastic Asphalt (SMA), Rubberized Open Friction Course (rofc) and Rubberized Open Graded Course (rogc). The Close-Proximity Method and Controlled Pass-By Method were utilized to measure the noise generated by the tire-pavement interaction in the field. Sound measurements based on three vehicle sizes were recorded at the various vehicle speeds. In order to evaluate the noise reduction effectiveness of the proposed pavement designs, these were compared to the control which is a typical Ontario municipal asphalt pavement, Hot-Laid 3 mix (HL-3). A statistical analysis was also performed to evaluate if significant differences existed between mixes and within a mix with respect to the vehicle speeds andor vehicle sizes. Based on the preliminary field data results, the rofc and rogc have a similar noise reduction capability, and can achieve a maximum noise reduction of 3.3 dba as compared with HL-3. The SMA showed a slight reduction for medium and heavy vehicle sizes but did not demonstrate noise reduction for light vehicle.

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 3 INTRODUCTION Road traffic noise is becoming a major public concern. Researchers and engineers are investigating new pavement mix designs to reduce traffic noise, and thus reducing the need for noise barrier walls. Rubberized porous asphalt pavement mix has been shown to be a potential pavement type that can be used to reduce pavement noise in numerous agencies around the world. (1,2,3,4,5) Therefore, the Regional Municipality of Waterloo (RMOW), Ontario, Canada, in partnership with the University of Waterloo Centre for Pavement and Transportation Technology (CPATT), have also undertaken research into asphalt pavements to determine whether the noise resulting from the interaction between tires and pavement can be significantly reduced in the southern Ontario environment as compared to a typical pavement with a Hot-Laid 3 (HL-3) surface. In addition to rubberized asphalt pavement, this paper will also investigate the noise reducing capability of stone mastic asphalt (SMA) mix, which is mainly used for heavy traffic conditions. Two types of sound level measurements were used in this study: Close-Proximity Method and Controlled Pass-By Method. A statistical analysis will also be performed utilizing the noise measurement results to determine if there is significant difference between mixes or within a mix. Although the test sections were constructed in 2004 and only a limited amount of data is available at this time, it is expected that the RMOW and CPATT will continue to monitor performance over time. DESCRIPTION OF THE STUDIED PAVEMENTS Rubberized open friction course (rofc), rubberized open graded course (rogc) and stone mastic asphalt (SMA) were chosen to be the studied asphalt pavements. These three asphalt pavements will be compared with a typical commonly used mix in municipalities in Ontario, Hot-Laid 3 (HL-3) which consists of 15% of recycled asphalt pavement (RAP) in this study. The design thickness of each pavement type is 40 mm. Rubberized asphalt includes finely chopped rubber reclaimed from old tires, which is heated and blended with the liquid asphalt cement and coats and binds gravel or stone aggregates together, which is known as a wet process. rofc and rogc are also known as popcorn or porous asphalt which is a layer of asphalt that incorporates a skeleton of uniform aggregate size with a minimum of fines. (6) It is believed that they are able to provide significant noise reductions as compared to HL-3. In this study, the mix designs and gradations for the rofc and rogc are similar. The difference between these two mixes is related to the quality of the aggregates. The aggregates used in rofc are premium grade and must meet much higher standards for the other aggregate tests. The ones used in rogc are local aggregates and do not meet all the requirements to qualify as a premium mix. Since SMA is utilized for heavy traffic conditions and is gaining popularity in Canada and the United States in recent years, it was also included in this study. (7) SMA mixes also have a porous nature and may also have noise reduction capabilities as compared with HL-3. The SMA is a gap-graded asphalt mix which combines strong, coarse aggregate with a high asphalt content, but lacks medium-sized aggregates. (8) The aggregates used in SMA are premium grade and must meet much higher standards for the other aggregate tests. The asphalt binders for all studied pavements are PG 64-28. This means the binder is designed to comply with the performance criteria at an average 7-day maximum pavement

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 4 design temperature of 64 o C and at a minimum pavement design temperature of -28 o C. The gradation, aggregate quality, and basic mix design properties have been summarized in Table 1. TABLE 1 Job Mix Formula and Basic Mix Design Properties Job Mix Formula rofc rogc SMA HL-3 16.00 mm 100 100 100 100 3.20 mm 98.1 97.8 98.1 98.8 9.50 mm 64.5 71.7 63.2 84.9 4.75 mm 27.1 27.9 22.4 58.7 2.36 mm 20.6 16.5 19.3 50.2 1.18 mm 14.1 12.1 16.4 37.7 0.60 mm 11.1 8.8 14.7 22.7 0.30 mm 6.8 6.5 12.1 13.7 0.15 mm 3.6 5.1 9.9 7.7 0.075 mm 2.5 4.1 8.1 4.9 Type of Aggregate Premium Local Premium Local Asphalt Contents (%) 5.6 5.8 5.7 5.0 Air Voids (%) 6.9 8.6 3.9 3.7 VMA (%) 18.3 19.3 16.9 15.4 Stability (N) 7397 6234 NA 14111 Flow (0.25 mm) 13.0 16.1 NA 10.6 BRD 2.281 2.286 2.355 2.400 MRD 2.450 2.501 2.451 2.493 Sieve Size SITE DESCRIPTIONS Four types of asphalt pavement surface courses were placed in August 2004 on William-Hasting Line (Road 11), between Manser Road and Chalmers Forrest Road in Waterloo, Ontario, Canada as shown in Figures 1 and 2.

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 5 Waterloo FIGURE 1 Map of Waterloo, Ontario (Adopted from www.mapquest.ca). N SMA HL-3 rofc rogc FIGURE 2 Site map (Adopted from www.mapquest.ca).

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 6 This test site is located in a rural area surrounded by farmlands. Pavement surface courses were placed in an order of rofc, rogc, SMA, and HL-3 from east to west. The length of each type of pavement section is approximately 600 m. A 40-mm thick cold-in-place recycled layer was placed throughout the test section areas as the binder or base course. NOISE MEASUREMENT METHODS Traffic noise measurements were taken in September 2004 which is about one month after pavement placement. There were 13 testing vehicles used for the noise testing and they were divided into three categories: light (5 vehicles), medium (5 vehicles), and heavy (3 vehicles) vehicles, which are listed in Table 2. Each noise measurement consisted of a single test vehicle passing through the test site. The driver of the testing vehicle drove through the centerline of the test road at constant speeds of 60 kmh, 70 kmh, 80 kmh, and 90 kmh from east to west and then made a return trip. Thus two measurements for each speed were taken for each vehicle. However, since the two measurements are independent of each other, they can be considered as individual measurements. TABLE 2 Description of Testing Vehicles Vehicle Size Vehicles Type Specifications CarLight Truck 2 cars, 1 mini van, 2 light trucks 2-axle, 4 wheels, 9 passengers, 4500 kg Medium Truck 2 city buses, 3 city work trucks 2-axle, 6 wheels, 4500 to 12000 kg Heavy Truck 3 snow plow trucks 3-axle, design for hauling cargo, 12000 kg Two sound level measurement techniques were utilized in this analysis: the Close- Proximity Method (CPX) and the controlled Pass-By Method (PBM). These two methods are commonly used for measuring vehicle noise worldwide as authorized in the International Organization for Standardization (ISO). (9) During the noise testing, the entire area was closed. The total closure distance was 3.6 km on William Hasting Line from Manser Road to Chalmers Forrest Road. Also, the study site is located in a rural area and the road was closed during testing, therefore the ambient noise should be minimized and constant, and would not affect the traffic noise measurement. Close-Proximity Method The Close-Proximity Method (CPX) can give a good acoustic quality estimation of the homogeneous road surface over a long distance and under a variety of conditions. A typical CPX method uses at least two microphones mounted near a test vehicle tire to measure sound power or sound pressure which is generated directed from the interaction between the tire and pavement. (9,10) In this project, only one microphone was mounted on the test vehicle and was located approximately 45 cm (18 inches) to 50 cm (20 inches) away from the centre of the front or rear wheel as shown in Figure 3. The purpose of this set-up was to ensure that the CPX measurement would measure the direct noise generated from the interaction between vehicle tire and pavement and to avoid measuring the engine noise generated from the testing vehicle. The CPX method measured the sound level on each pavement section in terms of an equivalent sound pressure, L eq (in dba), which is the average sound level in a given period. Since each pavement section is approximately 600 m long, it is believed that the measured equivalent sound pressure will be a good representation of the sound level.

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 7 Pass-By Method FIGURE 3 Close-proximity Method microphone set-ups. The Pass-by Method (PBM) measures the sound as vehicles travel pass a stationary microphone which is usually located at more than 7.5 m away from the centerline of the road and more than 1.2 m above the pavement. Currently, there are two types of pass-by methods typically used in road traffic noise analysis: the statistical pass-by and the controlled pass-by methods. (9,10) The Statistical pass-by method records the sound level in the normal daily traffic stream. However, this technique requires very precise traffic data, such as number of vehicles pass-by, the vehicle sizes, and vehicle speeds during the measurement period. The controlled pass-by method is simpler than the statistical method; it requires one vehicle or a small set of vehicles for the noise measurement. The sound level can be measured for different scenarios such as at constant speed or engine off and when the vehicle is in neutral. (9,10) In this study, the controlled pass-by method with a steady speed was used. Four pass-by monitoring stations (PBM) were set-up at the midway point of each asphalt pavement section as shown in Figure 4. FIGURE 4 Pass-by Method monitoring station set-ups. Each monitoring station was located 15 m away from the centreline of the road, 1.5 m above pavement, and was monitored by a technician. The maximum sound level (L max ) was

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 8 measured by the PBM. The maximum sound level (L max ) identifies the maximum sound levels produced by an event. The L max should be the sound level recorded at the shortest distance between test vehicle and the monitoring station in a single event. In this case, an event is defined as a test vehicle passing by. NOISE MEASUREMENT RESULTS Close-Proximity Method (CPX Method) Results The close-proximity method is used to measure the actualdirect noise generated between the tire and pavement. A graphical plot for the sound measurement versus vehicle speeds for each pavement type and vehicle size is shown in Figure 5. 98 Close-Proximity Method Average Sound Level of All Vehicle Sizes 96 Sound Level (Leq, dba) 94 92 90 88 86 84 60 kmh 70 kmh 80 kmh 90 kmh Vehicle Speed rofc-light rogc-light SMA-Light HL-3-Light rofc-medium rogc-medium SMA-Medium HL-3-Medium rofc-heavy rogc-heavy SMA-Heavy HL-3-Heavy FIGURE 5 CPX Method Average sound levels (L eq ) versus vehicle speeds. All four types of pavements show that when the vehicle speed or size increases, the sound level increases. The noise measurement range (the difference between the highest and the lowest noise measurement) for all speed limits are approximately 8 dba. The sound measurement results show that HL-3 has the highest sound level measurement for the heavy and medium vehicle categories for all four of the test speeds. It is noteworthy that the SMA has the highest sound level measurement for the light vehicle category for all speeds. Both the rofc and rogc have the lowest sound level measurement in all four vehicle speeds and three vehicle sizes. Also, the sound levels of rofc and rogc are very similar and it is expected they would have a similar

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 9 noise reduction result since the mix design in rofc and rogc are the very close except for the quality of the aggregate. The amount of noise reduction with respect to HL-3 is shown in Figure 6 that is presented by all vehicles, different vehicle speeds, and different vehicle sizes in a particular speed. Close-Proximity Method Average Sound Level Reduction for Different Vehicle Speeds and Sizes as compared to the Control, HL-3 4 Sound Level (dba) Reduced Gained 3 2 1 0-1 -2-3 -4-1.8-2.0 0.0 All Vehicle -1.4-1.7 0.5 60-0.4-0.6 2.3 60 - Light -1.8-2.4-0.7 60 - Medium -2.5-2.5-0.7 60 - Heavy -1.7-1.8 0.3 70 rofc rogc SMA -0.7-0.9 1.8-2.9-2.8-1.0-0.3-0.2 1.6-1.8-2.0-2.2-2.3-1.0-1.2 70 - Light 70 - Medium 70 - Heavy 80 80 - Light -3.3-3.3-1.6 80 - Medium -2.3-2.4-0.7 80 - Heavy -1.9-2.3-0.5 90-1.2-1.8 1.0 90 - Light -2.7-3.1-1.6 90 - Medium -1.8-1.9-1.1 90 - Heavy Vehicle Speed kmh - Vehicle Size FIGURE 6 CPX Method Average sound level reduction for different vehicle sizes as compared to the control, HL-3. In this initial study, the SMA showed the least amount of noise reduction and showed differences between the categories. The CPX test shows that SMA did not reduce the direct sound level as compared with HL-3 in the light vehicle category, at the speeds of 60 kmh and 70 kmh. Both the rofc and rogc pavements showed the greatest amount of direct noise reduction as compared with the HL-3. The highest noise reduction for rofc and rogc pavements is approximately 3.3 dba as compared with HL-3 in medium vehicle size at the speed of 80 kmh. The CPX results has also shown that all three kinds of pavements in this study can reduce noise for the medium sized vehicle than the light and heavy sizes of vehicles when the vehicle speed is over 70 kmh. In the overall performance (i.e. in terms of speed limit only), the highest noise reduction for rofc is 2.2 dba at the speed of 80 kmh and 2.3 dba for rogc at the vehicle speeds over 80 kmh. SMA only reduces noise by 0.5 dba when the vehicle speeds are at 90 kmh as compared with the HL-3. Pass-By Method (PBM) Results The PBM tests measured the direct noise and in-direct noise (reflected noise) generated by the interaction between the tire and pavements at 15 m distance from the vehicle. The noise measurement results are presented in terms of Maximum Sound Level (L max ), which is shown in Figure 7.

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 10 80 Pass-By Method Maximum Sound Level (Lmax) of All Vehicles Sizes 76 Sound Level (Lmax, dba) 72 68 64 60 60 kmh 70 kmh 80 kmh 90 kmh Vehicle Speed kmh rofc-light rogc-light SMA-Light HL-3-Light rofc-medium rogc-medium SMA-Medium HL-3-Medium rofc-heavy rogc-heavy SMA-Heavy HL-3-Heavy FIGURE 7 PBM Average sound levels (L max ) versus vehicle speeds. The PBM sound measurements of all four types of pavements are similar in terms of performance when compared to the CPX measurements. As the vehicle speed or size increases, the sound level increases. The noise measurement range for the vehicle speeds is approximately 10 dba to 12 dba. It also shows that SMA has the highest sound level measurement in light vehicle size for all testing speeds. HL-3 has the highest sound level measurement in medium and heavy vehicle sizes for testing speeds of above 70 kmh. However at the speed of 60 kmh in the medium and heavy sized vehicles, the sound measurement level of SMA and HL-3 are similar and are the highest among all pavement mixes. Both the rofc and rogc have the lowest sound level measurement in all three vehicle sizes. The amount of noise reduction with respect to HL-3 is shown in Figure 8 which is presented by all vehicle, different vehicle speeds, and different vehicle sizes in a particular speed.

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 11 4 Pass-By Method Maximum Sound Level Reduction for Different Vehicle Speeds and Sizes as compared to the Control, HL-3 3 rofc rogc SMA Sound Level (dba) Reduced Gained 2 1 0-1 -2-3 -1.0-1.7 0.2-0.6-1.2 0.7-0.1-0.5 2.1-0.3-1.9-0.6-1.9-1.6 0.1-0.8-1.3 0.4-0.5-0.5 1.7-1.2-2.4-1.2-0.8-1.7-0.3-1.4-2.2 0.0-0.6-1.6 1.4-1.8-2.8-1.6-2.5-2.4-0.5-1.5-2.3-0.1-1.2-2.3 1.2-2.4-2.8-1.4-1.2-1.8-0.7-4 All Vehicle 60 60 - Light 60 - Medium 60 - Heavy 70 70 - Light 70 - Medium 70 - Heavy 80 80 - Light 80 - Medium 80 - Heavy 90 90 - Light 90 - Medium 90 - Heavy Vehicle Speed kmh - Vehicle Size FIGURE 8 PBM Maximum sound level reduction in different vehicle sizes as compared to the control, HL-3. The pass-by noise reduction results illustrate that rofc and rogc provide a significant amount of noise reduction as compared with HL-3 in all situations. SMA does not reduce noise as compared to HL-3 for various comparisons, such as for vehicle speeds of 60 kmh and 70 kmh, and for light vehicle sizes at any speeds, which is similar as the CPX method results. It is observed that rogc performed the best in most of the comparisons. Although both rofc and rogc can reduce noise as compared with HL-3, rogc performed better than rofc in most of the cases except for the performance in heavy sized vehicle at 60 kmh and 80 kmh. In terms of vehicle sizes, rogc can reduce a large amount of noise for the medium sized vehicles for all vehicle speeds. The highest amount of noise reduction in rogc is 2.8 dba for the medium sized vehicle at the speeds of 80 kmh and 90 kmh. The highest noise reduction value observed in rofc is 2.5 dba for the heavy sized vehicles at the speed of 80 kmh. The lowest noise reduction values for rofc and rogc are approximately 0.1 dba and 0.5 dba, respectively. For the SMA pavements, the largest amount of noise reduction as compared with HL-3 is 1.6 dba at the medium sized vehicles at 80 kmh. In terms of the speed limit, rofc and rogc performed better when the vehicle speed is over 80 kmh. The highest noise reductions for rofc and rogc at the speed of 90 kmh are 2.2 dba and 2.3 dba, respectively. STATISTICAL ANALYSIS Figures 5 to 8 show that most of the pavement types result in similar noise levels in terms of vehicle sizes andor vehicle speeds. Therefore, a statistical analysis is required to evaluate if there are significant differences between mixes or within mixes. The statistical analysis using the t-distribution hypothesis testing method was carried out to examine if a significant difference existed between the asphalt mixes and within a particular asphalt mix in terms of vehicle speeds andor sizes. The paired comparison method was used to compare the similarities between pavements by performing a t-distribution hypothesis test on the difference between two

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 12 pavement mixes. Using this statistical analysis technique, it is possible to eliminate the variability present among vehicles when comparing the sound measurement between mixes. (11) This is very important since no two vehicle tires and weights are the same. A typical t- distribution hypothesis test was utilized to compare the similarity within pavement by comparing two sets of data in terms of vehicle sizes andor speeds in the same mix. The null hypothesis for testing whether there is significant difference between pavement mixes and within a pavement mix are as follow. Between Mixes (Blocking Paired Comparison Method) H o : µ D = 0, there is no significant difference between mixes H 1 : µ D 0, there is significant difference between mixes where µ D is the average value of difference between each pair of data point (A B) in a particular vehicle speed and size in two different kind of mixes. (e.g. D = A B, where A and B are the noise measurements of the same vehicle on two different pavements.) Within a Mix H o : µ 1 = µ 2, there is no significant difference within a mix H 1 : µ 1 µ 2, there is significant difference within a mix where µ n is the average of the set of the data point in a particular vehicle speed and size in a particular mix. The t-distribution hypothesis testing method was used for both analyses, between mixes and within a mix. The null hypothesis was rejected if the t value calculated was greater than the t critical at 95% confidence level. Four asphalt pavement mixes were studied in this project: rofc, rogc, SMA, and HL-3, therefore six comparisons in between mixes in terms of vehicle speed andor size were analyzed. These six comparisons are rofcrogc, rofcsma, rofchl-3, rogcsma, rogchl-3, and SMAHL-3. Comparison within mix was also analyzed in terms of vehicle speed andor size. A summary of the statistical analysis between mixes for the two test methods is shown in Table 3. A statistical analysis of the CPX results reveals there is no significant difference between rofc and rogc in terms of vehicle size andor vehicle speeds at a 95% confidence interval, except for the comparison on all medium and light sized vehicles, speeds of 60 kmh and 90 kmh, and the light vehicle size at the speed of 90 kmh. It is expected there is no significant difference between rofc and rogc in most of the comparisons because their gradations and basic mix design properties are similar. The statistical analysis of the CPX results also revealed that most of the other paired pavements, rofcsma, rofchl-3, rogcsma, and rogchl-3, are statistically significant different. However, it also shows that SMA and HL-3 are not significant different in terms of all vehicle speed, but they are significantly different in terms of measurements taken at various vehicle sizes. The statistical analysis results for the PBM reveals that some of the pavements have no significant difference in terms of vehicle sizes andor vehicle speeds. Especially for the comparison between rofc and rogc, there is no significant differences when the speed limit is 70 kmh, heavy sized vehicle, any vehicle sizes at the vehicle speed of 70 kmh, light and heavy vehicle sizes in the speeds of 60 kmh and 80 kmh, and medium and heavy vehicle size in the

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 13 speed of 90 kmh. For the comparison between rofc and SMA, it shows that rofc and SMA are statistically different for all vehicle speeds studied, but there is no significant difference in heavy sized vehicle and in all medium and heavy sized vehicles at any vehicle speeds. The statistical comparison results between rogc and SMA are similar as the comparison between rofc and SMA. However, it shows that there is significant difference in heavy sized vehicles and heavy sized vehicle at 70 kmh and 80 kmh. The statistical results show that rofc and HL- 3 have no significant difference at the light and heavy sized vehicles when the vehicle speed is below 90 kmh and when the vehicle speed is 60 kmh. However, rogc and HL-3 are significantly different except at the light sized vehicle at 60 kmh and 70 kmh. Also, the PBM statistical results reveal that the SMA and HL-3 are more similar than the CPX results.

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 14 TABLE 3 Summary of Statistical Analysis between Mixes CPX and PBM Methods CPX: L eq PBM: L max Pavement Types All Vehicle rofc rogc rofc SMA rofc HL-3 rogc SMA rogc HL-3 SMA HL-3 rofc rogc rofc SMA rofc HL-3 rogc SMA rogc HL-3 Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes No Heavy No Yes Yes Yes Yes Yes No Yes Yes Yes Yes No Medium Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Light Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes 60 Yes Yes Yes Yes Yes No Yes Yes No Yes Yes Yes 60-L No Yes No Yes Yes No No Yes No Yes No No 60-M No Yes Yes Yes Yes No Yes No Yes No Yes No 60-H No Yes No No No No No No No No Yes No 70 No Yes Yes Yes Yes No No Yes Yes Yes Yes No 70-L No Yes Yes Yes Yes Yes No Yes No Yes No Yes 70-M No Yes Yes Yes Yes No No No Yes No Yes No 70-H No Yes Yes Yes Yes No No No No Yes Yes No 80 No Yes Yes Yes Yes No Yes Yes Yes Yes Yes No 80-L No Yes Yes Yes Yes Yes No Yes No Yes Yes Yes 80-M No Yes Yes Yes Yes Yes Yes No Yes No Yes Yes 80-H No Yes Yes Yes Yes No No No No Yes Yes No 90 Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes No 90-L Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes 90-M No Yes Yes Yes Yes Yes No No Yes No Yes No 90-H No Yes Yes Yes Yes Yes No No Yes No Yes No Yes = Statistically Significant Difference No = No Statistically Significant Difference Table 4 is a summary of the statistical analysis within pavement (rofc, rogc, SMA, and HL-3). Four types of pavements have the same statistical results in terms of the comparison between vehicle speeds andor vehicle sizes within its pavement type. In CPX statistical analysis results, all the pavements have showed that there are no significant differences in the comparison between 7080 kmh, 8090 kmh and all comparison between medium and heavy vehicles in all speeds, except for SMA, which has significant difference between medium and heavy vehicles at the speed of 80 kmh. The PBM statistical analysis results are similar to the CPX statistical analysis results especially for the comparison between medium and heavy sized vehicles at any speeds. CPX statistical result shows that there are no significant differences in the comparison between 6070 kmh, 7080 kmh, and 8090 kmh. Also, the PBM statistical analysis has also shown that there are no significant differences in the comparison of 6080 kmh in rofc and rogc, 7090 kmh in rogc and SMA. SMA HL-3

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 15 CONCLUSIONS TABLE 4 Summary of Statistical Analysis within Mixes CPX and PBM Method CPX: L eq PBM: L max Pavement rofc rogc SMA HL-3 rofc rogc SMA HL-3 6070 Yes Yes Yes Yes No No No No 6080 Yes Yes Yes Yes No No Yes Yes 6090 Yes Yes Yes Yes Yes Yes Yes Yes 7080 No No No No No No No No 7090 Yes Yes Yes Yes Yes No Yes No 8090 No No No No No No No No 60L-60M Yes Yes Yes Yes Yes Yes Yes Yes 60L-60H Yes Yes Yes Yes Yes Yes Yes Yes 60M-60H No No No No No No No No 70L-70M Yes Yes Yes Yes Yes Yes Yes Yes 70L-70H Yes Yes Yes Yes Yes Yes Yes Yes 70M-70H No No No No No No No No 80L-80M Yes Yes Yes Yes Yes Yes Yes Yes 80L-80H Yes Yes Yes Yes Yes Yes Yes Yes 80M-80H No No Yes No No No Yes No 90L-90M Yes Yes Yes Yes Yes Yes Yes Yes 90L-90H Yes Yes Yes Yes Yes Yes Yes Yes 90M-90H No No No No No No No No Yes = Statistically Significant Difference No = No Statistically Significant Difference The Regional Municipality of Waterloo, Ontario, Canada, in collaboration with the University of Waterloo, Centre for Pavement and Transportation Technology have measured road traffic noise levels on three types of specially designed noise reducing pavements, Rubberized Open Friction Course Asphalt Pavements (rofc), Rubberized Open Graded Friction Course (rogc), and Stone Mastic Asphalt Pavement (SMA). The two types of open friction course asphalt pavements have a similar mix design, but different in the quality of the aggregate. One type contains premium aggregates (rofc) and the other contains local aggregates (rogc). SMA is a pavement for a heavy traffic area and which are also commonly used in the North America. The purpose of noise testing is to determine whether these three kinds of pavements provide significant noise reductions compared to a typical pavement with a Hot-Laid 3 (HL-3) surface. Two types of noise measurements were taken in this study: Close-Proximity Method (CPX) and controlled Pass-By Method (CPM). It is important to note that these findings are preliminary as only one set of measurement has been taken at this point. It was also found that vehicle noise increases when the vehicle speed or size increases for both test methods. The SMA pavement type did not provide any noise reductions for light sized vehicles and for vehicle speeds of 60 kmh and 70 kmh in both test methods. rofc and rogc provided the highest amount of noise reduction in both testing methods. The highest noise reduction amount for rofc and rogc pavement was 3.3 dba in the CPX results. For the PBM

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 16 results, the highest noise reduction amounts were 2.5 dba and 2.8 dba for rofc and rogc pavement, respectively. The statistical analysis in both measurement methods shows that there was no significant difference between rofc and rogc in most of the vehicle sizes andor speeds comparisons. It also had the same results between SMA and HL-3; there was no significant difference in some comparisons between vehicle sizes andor speeds. In the statistical analysis of within mix, all pavements had no significant difference when comparing the medium and heavy vehicles at any vehicles speeds, except at the speed of 80 kmh in SMA. Also, all pavements had no significant difference when comparing the vehicle speed of 80 kmh to 70 kmh and 90 kmh in the both measuring methods. However, all pavements also had no significant difference in PBM results when comparing the vehicle speeds of 60 kmh to 70 kmh, and when rofc and rogc had no significant difference at the comparison between 6080 kmh and when rogc and HL-3 had no significant difference at the comparison between 7090 kmh. RECOMMENDATIONS This paper presents initial findings on the effectiveness of several noise reducing pavements based on field measurements, it is recommended that additional sound level measurements be conducted in the future to monitor the pavement acoustical performance. If possible, performing repeated measurements for the two test methods (Close Proximity Method and Controlled Pass- By Method) to assess the measurement accuracy of the pavement noise reduction capability. Also, a comparison of the noise measurement results obtained from the two test methods should be performed and measurements of pavement acoustical absorption should also be conducted. A life cycle cost for each pavement should also be performed for noise reducing pavement selection. Also, a further analysis of the correlation between mix design and sound level measurement should be performed which may be used to act as a noise reduction prediction model. ACKNOWLEDGEMENTS The authors would like to acknowledge the Regional Municipality of Waterloo in Ontario and the Centre for Pavement and Transportation Technology through the Canada Foundation for Innovation and the Ontario Innovation Trust Fund for providing funding for this project.

Fiona Leung, Susan Tighe, Scott Penton, Gary MacDonald 17 REFERENCES [1] British Columbia Ministry of Transportation and Highways, Open Graded Asphalt Quiet Pavement : Assessment of Traffic Noise Reduction Performance, Phase 1, November 1995, BC, Canada, 1996 [2] Scofield, L., Donavan, P., The Road To Quiet Neighborhoods In Arizona, Presented at 85 th Annual Meeting of the Transportation Research Board, Washington, D.C., 2005 [3] Acoustical Analysis Associates Incorporated, AAAI Report 1272: Asphalt Rubber Overlay Noise Study Update, Prepared for County of Sacramento Public Works Agency Department of Transportation, Simi Valley, CA, 2002 [4] Kandhal, P. S., How Asphalt Pavement Mitigate Tire-Pavement Noise, In Better Roads, Special Asphalt Section, 2003 [5] McMillan, C., Kwan, A., Donovan, H., Enslen, P., Horton, B., Current Asphalt Rubber Developments in Alberta, Prepared at 2003 Annual Conference of the Transportation Association of Canada, St. John s, Newfoundland and Labrador, 2003 [6] Kuennen, T., Open-Graded Mixes: Better the Second Time Around. American City and County,1996, http:apwa.americancityandcounty.comargovernment_opengraded_mixes_better, Accessed November 20, 2004. [7] Hass, R., Pavement Design and Management Guide, TAC, Ottawa, Ontario, 1997 [8] SMA Proves Its Long-Term Durability, Asphalt Pavement Alliance, http:www.asphaltalliance.comuploadsma%20proves%20its%20long- Term%20Durability.pdf#search='Stone Matrix Asphalt strong course aggregate', Accessed July 11, 2005. [9] Sanberg, U., Ejsmont, J.A., TireRoad Noise Reference Book, First Edition, Poland, 2002. pp.307-310 [10] McDaniel, R., Thornton, W.; Field Evaluation of a Porous Friction Course for Noise Control, Prepared at 85 th Annual Meeting of the Transportation Research Board, Washington, D.C., 2005. [11] Duever, T., ChE 622 Statistics in Engineering Course Notes, Waterloo, ON, Canada, 2002.