CASE STUDY OF TYRE NOISE: ASSESSMENT AND COMPARISON OF DIFFERENT ROAD SURFACES

CASE STUDY OF TYRE NOISE: ASSESSMENT AND COMPARISON OF DIFFERENT ROAD SURFACES W Mior & M H F de Salis Vipac Engineers and Scientists Ltd Unit E1-B Centrecourt, 25 Paul Street Nth North Ryde, NSW, 2113. Email: wanm@vipac.com.au 1 INTRODUCTION This paper describes work involved with an acoustic impact assessment of the conversion of a section of road from a 50 km/h thoroughfare surfaced with dense grade asphalt to a lower limiting speed shared zone surfaced with textured porphyry setts. Traffic in the study area largely consists of light-weight vehicles. The study addressed concerns that noise generated from the new shared zone might be higher than previously experienced. The aim of the study was to determine comparative tyre/road noise levels generated by lightweight vehicles travelling over asphalt and porphyry setts surfaces, the latter being similar to a granite setts surface. Subsequently this information was used to provide guidance in setting a new speed limit for the porphyry surface so that tyre/road noise from this section would not exceed that from the original asphalt surface. 2 REVIEW OF TYRE NOISE MEASUREMENT TECHNIQUES The relationship between tyre noise and road surface texture is appreciably complex as it depends on many factors including type of vehicle, type of tyre, vehicle speed, road surface material and weather conditions 1. European researchers commonly use the Statistical Pass-by (SPB) and Close-Proximity (CPX) methods to study tyre/road noise. The first part of ISO 11819 outlines the SPB method and the second part describes the CPX method for tyre/road noise measurement. The SPB method measures a statistically significant number of individual vehicle pass-by noise levels (L Amax ) at a 7.5m horizontal distance from the centre of the vehicle trajectory, 1.2m above the ground. The Statistical Pass-By Index (SPBI) can be calculated from data collected using the SPB method to indicate the influence of the road surface on traffic noise. Two factors are taken into consideration when calculating the SPBI: the mix of vehicles including light-weight, dual-axle and multi-axle vehicles and the speed of vehicles i.e. either low (45 64 km/h), medium (65 99 km/h) or high speed (100 km/h and above). The SPBI

of different test surfaces can then be compared with a reference surface for road noise and surface ranking purposes 2. The SPB method is only reliable for vehicles travelling at or above a minimum steady speed. Tyre/road noise dominates vehicle noise emissions at speeds of 45 km/h and higher for lightweight vehicles and 80 km/h and higher for heavy vehicles. At lower speeds, acceleration and deceleration can lead to engine and drive-train noise which outweighs tyre/road noise 2. The Close-Proximity (CPX) or trailer method has also been used in European tyre/road noise studies to test different road conditions i.e. any road surface, at any speed and with any tyres. Unlike the SPB method, the CPX method involves specific measurements of tyre noise using a microphone mounted approximately 200mm to 400mm away from the test tyre sidewall 3. The TRITON test vehicle was specifically designed by Bruel & Kjaer for tyre/road noise investigation using the CPX method. It has a specially designed body that incorporates an anechoic chamber enclosing the test tyre and an array of microphones. The chamber isolates the microphones from external noise sources such as the truck engine, other tyres or other vehicles 3. The CPX method does not accurately represent the affect of tyre/road noise at the road side since it does not include for the effect of the car body or effect of engine noise of the car body or the car s surroundings 4. However, the evolution of road-surface properties over time and space can be recorded using this method 4. AS 2240-1979 5 details procedures for the measurement of sound emitted by motor vehicles. Procedures in AS 2240 are defined such that repeatable tests of overall vehicle noise can be undertaken for comparison with guideline noise emission criteria. 3 METHODOLOGY In the current study a measurement was devised for comparative testing of noise generated from the tyres of a test vehicle travelling over two separate road surfaces, namely asphalt and porphyry setts. This measurement procedure draws on the SPB and CPX methods and varies some of the general methodologies of AS2240 to achieve maximum measured tyre noise level received by the microphone. The devised procedure was used to investigate tyre/road noise generated by vehicles travelling over a porphyry setts surface at very low speeds i.e 10, 20 and 30 km/h indicative of the proposed traffic conditions at the shared zone. Measurements were repeated over the existing asphalt surfaces at 30 and 50 km/h indicative of the current traffic conditions at the site. Discounting acceleration and deceleration effects, it seems reasonable to expect that overall vehicle noise will increase with vehicle speed. On smooth road surfaces engine noise is generally accepted as being the dominant factor in vehicle noise generation at speeds below about 45 km/h and tyre noise tends to become the dominant factor at speeds in excess of 45 km/h 2. However at speeds below 45 km/h, we assumed that for a fixed speed of our test

vehicle the engine noise would remain constant and at a low noise level over different road surface types as the test vehicle coasts over the road surface during measurement. In light of the above where changes in measured car pass-by noise levels were recorded for different road surfaces at a constant vehicle speed, we have assumed that this was due to changes in tyre noise. 3.1. Measurement Environment 3.1.1. Existing Asphalt Road Surface Measurements of tyre noise were initially undertaken on the existing road surface at the planned shared traffic zone at Test Site No. 1 (asphalt road surface). The existing speed limit in the area was 50 km/h and as such measurements of tyre noise were taken for vehicle passbys at 50 km/h and also at 30 km/h for reference. The test vehicle was a Holden Acclaim 3.8L station wagon owned by Vipac at the time of the tests. The vehicle had 50-70% worn tyres. 3.1.2. Porphyry Setts Test Surface Measurements were repeated at Test Site No. 2, which had a section of road with a raised surface of porphyry setts approximately 15 metres long, similar to the Council s planned shared traffic zone at Test Site No. 1. As it was proposed to impose a low speed limit over the porphyry setts surface in the planned shared traffic zone, measurements were taken for test vehicle pass-bys at 10, 20 and 30 km/h for comparison with the asphalt surface tests. Measurements were undertaken with an integrating type 1 sound level meter incorporating a free-field microphone. The measurement system was field calibrated before and after the measurements. 3.2. Measurement Indices Measurement indices recorded were in line with AS2240 which states that measurements of pass-by vehicle noise should be measured using a sound level meter set to A weighting and fast response. In fast response mode the sound level meter samples the noise climate every 0.25 seconds. The recorded sound level reading is the highest fast response noise level obtained during the pass-by measurement, i.e. the L Amax level. The L Amax value would be expected to occur when the vehicle passes closest to the microphone so the ratio of tyregenerated noise levels to underlying ambient noise levels is maximised at that instant. Use of L Amax as the noise level indicator during the measurement rather than an equivalent L Aeq averaged over a number of fast response samples reduces the likelihood of underlying ambient noise influencing the measurement.

Representative measurements of ambient noise were recorded to ensure that extraneous noise levels were low enough to have a negligible effect on the measurements. In this case L Aeq equivalent noise levels recorded separately over an ambient noise measurement period were taken to represent the underlying ambient noise existing during pass-by measurements. 3.3. Measurement Orientation Measurements were undertaken with the microphone elevated 0.5m from the ground pointing towards the vehicle closest pass-by point and at 2m from the vehicles centreline of passage. This represents a departure from AS2240 which recommends that the microphone be mounted at a height of 1.2m, 7.5m from the centreline of the vehicle s path. Departure from the standard test procedure at this stage is considered acceptable as noise levels are being measured for comparison using a specific repeatable test procedure. Positioning of the microphone closer to the test vehicle helped to increase signal levels at the microphone and positioning closer to the ground was aimed at focussing on tyre noise. The asphalt surface test site had buildings on the opposing side of the road to the measurement position at a distance of approximately 15 m. Building reflections were assumed to be shielded by the car at the point of maximum noise level. The close proximity of the microphone to the passing vehicle ensured that the magnitude of reflected noise was as small as possible in comparison to the direct noise measured at the microphone. Measurements at the porphyry setts test site were far enough from direct reflecting surfaces to be able to discount the possibility of reflection effects interfering with measurement results. 3.4. Number of Measurements AS2240 notes that measurements should be repeated until three consecutive measurements agree to within 2 db. 3.5. Vehicle Velocity Timed records of vehicle pass-bys over a measured distance indicated that the speedometer of the test vehicle was accurate to within 2 km/h at speeds of 20, 30 and 50 km/h. Speedometer readings at vehicle speeds of 10 km/h and below were found to be inaccurate. As such vehicle speeds for tests at 10 km/h were checked by the time taken to traverse the test strip (of known length).

4 EXPERIMENTAL RESULTS The measurement procedure was undertaken as described in Section 3. The before and after calibration levels of the measurement system agreed to within 0.2 db tolerance. In all cases the underlying ambient L Aeq noise levels were more than 10 db below the measurement L Amax levels measured during vehicle pass-bys. This indicated that statistically, the vehicle pass-by measurement data would be unaffected by extraneous noise. It was noted that tyre noise was dominant as the car coasted past the measurement position. 4.1. Vehicle Pass-by Tests on an Asphalt Road Surface Table 1 shows the results from vehicle pass-by noise levels measurement on the asphalt road test site at the planned shared traffic zone. Ambient noise levels recorded at the site are shown in Table 2. Vehicle Pass-by Speed Test 1 Test 2 Test 3 Average L Amax db(a) 30 km/h 70.5 71 70 70.5 50 km/h 77 77 76.5 76.8 Table 1 L Amax Noise Levels Recorded During Pass-By Tests for Test Vehicle travelling over Standard Asphalt Road at Test Site No. 1 Indicator Measured L Aeq L Amax L A10 L A90 Ambient Levels db(a) 54 57.5 55 52 Table 2 Ambient Noise Measurements at Test Site No. 1 Averaged spectra for the vehicle pass-bys associated with the results in Table 1 and L Aeq ambient noise in Table 2 are shown in Figure 1 in db(a) and Figure 2 in db(linear). Figure 1 Spectra of L Amax Pass-by Noise and Ambient L Aeq in db(a) at Standard Asphalt Road Test Site No. 1

Figure 2 Spectra of L max Pass-by Noise and Ambient L eq in db(linear) at Standard Asphalt Road Test Site No. 1 4.2. Vehicle Pass-by Tests on a Porphyry Setts Road Surface Table 3 shows the results from vehicle pass-by noise levels measurement at the porphyry setts test site. Ambient noise levels recorded at the site are shown in Table 4. Vehicle Pass-by Speed Test 1 Test 2 Test 3 Average L Amax db(a) 10 km/h 72.5 71.5 72 72 20 km/h 76.5 74.5 76 75.7 30 km/h 80 80.5 79.5 80 Table 3 L Amax Noise Levels Recorded During Pass-By Tests for Test Vehicle travelling over Porphyry Setts at Test site No. 2 Indicator Measured L Aeq L Amax L A10 L A90 Ambient Levels db(a) - 1 60.5 66 62.5 57 Ambient Levels db(a) - 2 57 64.5 59 54 Table 4 Ambient Noise Measurements at Test Site No. 2 Averaged spectra for the vehicle pass-bys associated with the results in Table 3 and L Aeq ambient noise levels in Table 4 are shown in Figures 3 and 4 in db(a) and db(linear) respectively.

Figure 3 Spectra of L Amax Pass-by Noise Levels and Ambient L Aeq in db(a) at Porphyry Setts Surface at Test Site No. 2 Figure 4 Spectra of L max Pass-by Noise Levels and Ambient L eq in db(linear) at Porphyry Setts Surface at Test Site No. 2 4.3. Comparison of Vehicle Pass-by Measurements at 30 km/h Figure 5 shows the linear weighting plot of vehicle pass-by spectra at both the asphalt and porphyry setts test sites.

Figure 5 Spectra of L max Pass-by Noise Levels at 30 km/h in db(linear) at Asphalt and Porphyry Setts Test Sites 5 DISCUSSION OF MEASURED TYRE NOISE CHARACTERISTICS Tables 1 and 3 show an increase in measured noise levels as the vehicle speed increases. Ambient noise levels in Tables 2 and 4 are at least 10 db(a) below any of the vehicle pass-by measurement levels. The ambient L Amax levels came within 10 db(a) of some lower speed measurements but would have been statistically unlikely to repeat simultaneously with the vehicle L Amax during pass-by measurements. This is underlined by the consistency of measurements in Tables 1 and 3. Pass-by spectra were originally recorded to determine whether the pass-by measurements would be considered as tonal. While the measurements were subsequently not considered as being tonal, the spectra are of interest for comparison with accepted tyre noise generation mechanism theory. The consistency of shape in the dominant frequency range of pass-by spectra was also helpful as it indicated that pass-by measurements were unaffected by random extraneous noise. In the Figure 1, A weighted spectra for the smooth asphalt surface, a general peak in the 1 khz to 2 khz frequency range is apparent for both spectra. In the Figure 3, A weighted spectra for the porphyry setts surface, all spectra peak at a lower frequency of 1 khz. However, peaks for either surface are not always so apparent in the linear spectra of Figures 2 and 4. There is no strongly apparent relationship between peak frequencies and vehicle speed for either surface. Tyre noise generation mechanism varies depending on the road surface texture pattern. Noise from tyre interaction with smooth road surfaces (random texture pattern) are most often caused by air pumping effects which tend to produce sound waves in the frequency range 1 to 3 khz. This corresponds with the peak of the spectrum in Figures 1 and 2. The horn effect, which is associated with the air pumping effect, arises from an amplification of noise generated at or near the contact patch which tends to focus the sound. The largest

amplifications have been reported in the 2 khz region 1. This corresponds with the peak frequency for 50 km/h in Figure 1. Another sound generation mechanism is the snap out effect where vibration is induced by the tyre tread block snapping back to its undeflected rolling radius and tyre thread impact with the texture pattern. This is often the main noise generating mechanism for tyre interaction with rough road surfaces such as porphyry setts. In this case low frequency vibrational modes of the tyre become dominant as a noise radiating mechanism and the tyre acts as a low pass filter with cut-off frequency generally around 1 khz 1. This cut-off is apparent in Figures 3 and 4. Air pumping effects generally reduce on uneven surfaces due to leakage and irregular cavity volume. Consequently a noticeable reduction of level in the air pumping dominant frequency range of 1 to 3 khz is apparent in Figures 3 and 4 compared to Figures 1 and 2. In Figure 5, it can be seen that there is more energy in the low to middle frequency range (200 Hz to 1 khz) of the pass-by spectrum recorded at the porphyry setts surface compared to asphalt. This concurs with the low frequency snap out noise generation theory for rough surfaces discussed above. The porphyry setts were spaced at 0.1m giving a pass-by frequency of around 80 Hz at 30 km/h which concurs with the low frequency peak in this region in Figure 5. No concurrent peaks occur for lower speeds in Figures 3 and 4, although pass-by frequencies may bisect tyre mode frequencies at these lower speeds causing suppression of peaks. However, there are also strong low frequency peaks in Figures 1 and 2 for asphalt which are not so readily attributable to tyre noise, suggesting that perhaps such peaks are caused by some other mechanism than periodic surface irregularity. 6 CONCLUSIONS As anticipated in the Methodology section, for a given vehicle velocity the measured noise is louder over the porphyry setts surface than the asphalt surface. It was observed by the operator that noise from the vehicle was dominated by tyre noise. Spectral shapes recorded generally concur with the accepted theory on tyre noise generation mechanisms. Results indicated that noise levels generated due to tyre/road interaction on porphyry setts were appreciably higher than those generated on asphalt for similar speeds, with differences exceeding 9 db(a) at a speed of 30 km/h. A comparison of noise levels recorded for varying vehicle speeds at the different test sites is given in Table 5.

Vehicle Pass-by Speed and Road Surface Condition Average L Amax db(a) Vehicle Pass-by Speed and Road Surface Condition Average L Amax db(a) - - 10 km/h Porphyry Setts 72 - - 20 km/h Porphyry Setts 75.7 30 km/h Asphalt 70.6 30 km/h Porphyry Setts 80 50 km/h Asphalt 76.5 - - Table 5 Comparison of Measured Average L Amax levels on Different Surfaces at Different Vehicle Velocities A speed limit of 50 km/h is currently in place for the asphalt road surface at Test Site No. 1. Noise levels generated as a result of tyre interaction with a porphyry setts surface at 10 km/h and 20 km/h were found to be lower than noise levels generated as a result of tyre interaction with the asphalt surface at 50 km/h. Therefore we have concluded that noise generated from light-weight vehicles coasting over the planned porphyry setts surface at the proposed 10 km/h speed limit or even at 20 km/h will not adversely affect the residents closest to Test Site No. 1. Thus, we recommended that 10 km/h speed limit be imposed on the planned shared zone. We also recommended that any lead up gradients to the porphyry surface should be minimised to reduce bumping noise. REFERENCES 1. P M Nelson and S M Phillips (1997), Quieter Road Surfaces, Transport Research Laboratories (TRL) Annual Review, Berkshire UK. 2. ISO 11819-1 (1997), Acoustics Measurement of the influence of road surfaces on traffic noise -- Part 1: Statistical Pass-By method, ISO Standards, Geneva. 3. Bruel & Kjaer (Jan 2002), Pulse System measures Tyre/Road noise, www.sensorland.com/apppage045.html, UK. 4. V Meier (1995), Methods to Measure and Control the Acoustical Quality of Road Surfaces, International Congress on Acoustics (ICA) 95, Trondheim, Norway. 5. Australian Standard AS 2240 (1979), Methods of Measurement of the Sound Emitted by Motor Vehicles, Standards Association of Australia, Sydney.