To compare acoustical performances between different tires and pavements, various experimental techniques have been developed to measure tire/road noise.
These techniques can be broadly divided into two major categories, namely in-field measurements and laboratory measurements. The former involve full-scale test pavement sections, while the later are commonly conducted on test drums covered by pavement samples. Some techniques specify the vehicles and tires being used and others adopt statistical approaches to deal with the vehicle and tire variations. The widely accepted in-field measuring methods include the statistical pass-by method (SPB), controlled pass-by method (CPB), coast-by method (CB), close-proximity method (CPX) and on-board sound intensity method (OBSI). Most of these in-field
54
techniques have been standardized. The laboratory measurements vary from one to another without international specifications being established. They may use either a stationary wheel or a stationary pavement test set-up. These measurement methods are briefly introduced in this section, and their advantages and disadvantages are discussed.
2.2.3.1 Statistical Pass-By Method
Statistical pass-by method is a far field noise measurement approach utilizing a random sample of typical vehicles selected from the traffic stream, under constant or nearly constant speed conditions. This test is standardized by ISO 11819-1 (ISO, 1997a) in detail. The test site configuration is illustrated in Figure 2.17, in which a stationary microphone is located 7.5 m from the central line of the travel lane. The vehicles measured are those cruising-by the test site without disturbances from other vehicles. At the passage, the maximum A-weighted sound level is measured together with the vehicle speed. The vehicle is also classified into one of the preset categories.
After measuring at least 100 cars and 80 heavy vehicles, the measured sound levels are plotted against the measured speeds by vehicle category. Statistical methods are then used to assess the noise levels and derive the statistical pass-by Index (SPBI).
Statistical pass-by method is easy to conduct at a normal operation condition without traffic control. It provides a quick and direct evaluation on noise level near the roadway, and takes into account the vehicle and tire variation in the realistic traffic flow. SPB is particularly useful to compare the acoustic performance of several pavement sections constructed on the same road (Abbott and Phillips, 1996) and to conduct a "before-and-after" study when a new pavement treatment is performed on an existing road (Sandberg, 1997). The main disadvantage of SPB method lies in its loose definition of “normal” traffic condition. It assumes that the randomly selected vehicle sample is representative, however traffic composition depends on location and time and is not a constant. This makes it difficult to compare two pavements with
55 significantly different traffic compositions. SPB method measures the noise generated by the entire vehicle, including power unit noise and aerodynamic noise as well, other than tire/road noise. Another disadvantage of SPB method is the long measurement time it takes due to the large sample size collected. Moreover, as shared by all roadside noise measurement methods, the results are susceptible to environmental factors such as wind and temperature.
2.2.3.2 Controlled Pass-By Method
The test configuration of controlled pass-by method is similar to that of SPB method. Instead of a statistical traffic sampling, the vehicles and tires being tested in CPB method are prescribed. For each test location, a specific vehicle cruises by at a given speed with its engine running at normal condition and a maximum A-weighted noise level is measured at the roadside. Multiple test locations are necessary for each test section to derive the average noise level representing the full test section. The CPB method is specified by a national French standard (NF S 31-119-2), which makes use of two cars and four tire sets. It is used to survey the French road network, establishing a database on the acoustical properties of all the major French roadways (Sandberg and Ejsmont, 2002).
One of the primary advantages of the CPB method is that it can control most test variables, such as the vehicle, tire and test speed, making it easier to compare the acoustic properties of different pavement surfaces. CPB measurement can be performed much faster compared to SPB method because fewer runs are needed for generating reliable results. It also allows for passes at different speeds to derive the speed dependency of noise level. A drawback of CPB method is that it does not account for the variation in traffic composition and thus cannot estimate the perceived noise level at roadsides. CPB method requires test vehicles to be isolated from other traffic, and background noise has to be kept as low as possible. Therefore, potential problems may occur when conducting the measurements on highly trafficked roads.
56
This may lead to traffic control or closure of the road, causing inconvenience to road users. Similar to SPB measurement, the CPB method also measures the overall noise generated by the entire vehicle, rather than the separated tire/road noise.
2.2.3.3 Coast-By Method
Coast-by approach was developed to eliminate engine noise from the noise levels measured at roadsides. The test configuration is the same as CPB method, except that the test vehicle is coasting through the test area with the engine switched off and the transmission put in neutral. The vehicle is driven only by the inertia force and is assumed to emit only tire/road noise. During coasting, the vehicle will slightly slow down. This is usually neglected and the speed at the maximum noise level or the average speed is used in the data processing. The Coast-by method is standardized by ISO 13325 (ISO, 2003).
The principal advantage of CB method is that it eliminates the interference of power unit noise. The noise levels measured with this method can be directly used to estimate the tire/road noise emitted from individual vehicle at particular speeds. The CB method has disadvantages similar to the CPB approach and the CB measurement result is affected by the performance of test vehicle. Besides, narrow-band frequency analysis (narrower than 1/12 octave band) is inapplicable to all roadside measurement methods unless advanced Doppler compensation methods are employed. The Doppler effect has a negligible effect on one-third octave band frequency analysis.
2.2.3.4 Close Proximity Method
The close proximity method, previously named the trailer method, is a near field measurement technique specially designed to focus on the tire/road noise. This method is standardized by ISO 11819-2 (ISO, 2013). In the CPX measurements, a test tire is mounted on a trailer lined with the sound-absorbing material (see Figure 2.18). Two microphones are fixed near the leading and trailing edges of the contact
57 patch at 200 mm from tire side wall and 100 mm above road surface (see Figure 2.19).
The trailer is hooded by an enclosure to reduce the influence of wind noise and pulling vehicle noise. To control the uncertainties from test tire, standard reference tires are used in CPX measurements, as specified in ISO/TS 11819-3 (in preparation).
Some of the most widely used reference tires are illustrated in Figure 2.20. In the CPX tests, the average A-weighted sound pressure levels emitted from the tire-road interaction are measured over a specified distance, together with the vehicle speeds.
The close-proximity sound index (CPXI) is then derived at a reference speed and the spectrum analysis could be performed based on the time series data.
The CPX technique makes possible direct near-source measurements of tire/road noise. The primary advantage of this method is that measurements are fast and efficient. Once the test vehicle and devices are available, a large distance of pavement sections can be measured continuously within a single test run. Comparing to roadside methods, the CPX method is less susceptible to suffer from the variations of environmental conditions. Because of the enclosure installed around the test tire, no strict requirements on background noise are specified in the CPX method, and it can be conducted on roads with considerable traffic. Besides, narrow-band spectrum analysis is feasible with this method. However, there are still several disadvantages coming with the CPX method. It only measures a limited number of vehicles and tires, therefore it cannot represent the actual traffic composition, nor can it measure perceived noise in the far field. In CPX tests, the variation of measuring results due to different test vehicles and trailers is large. Higher investment is needed for the equipments and vehicles used in the tests, making it more expensive compared to the roadside methods.
2.2.3.5 On-Board Sound Intensity Method
The on-board sound intensity method was developed in the NCHRP Project 1-44 (Donavan and Lodico, 2009) to measure the tire/road noise at near field. It has
58
been standardized by AASHTO Specification TP 76-11 (AASHTO, 2011b) and ASTM WK26025 (ASTM, 2009e). The test configuration of OBSI technique is illustrated in Figure 2.21. The probes are mounted on one of the tires of a test vehicle instead of a trailer. A significant difference between the OBSI and CPX methods is that the OBSI method measures sound intensity instead of sound pressure (as in CPX method). Sound intensity measurement makes use of two phase-matched microphones arranged in a particular direction in one probe to measure sound intensity perpendicular to the travel direction. With this feature, sounds from other sources in other directions (i.e. wind noise, engine noise etc.) are attenuated, increasing the reliability in measuring tire/road noise. The standard reference test tire P225/60R16-97S (ASTM, 2008c) is adopted in the OBSI method and sound intensity probes are fixed at specific positions near the leading and trailing edges of contact patch. The OBSI method has advantages and disadvantages similar to the CPX method. OBSI method is found to better correlate mathmatically with roadside noise measurements (Donavan and Lodico, 2009).
2.2.3.6 Laboratory Measurement
Besides the above mentioned in-situ approaches, tire/road noise can also be measured in the laboratory. The laboratory measurement usually requires access to some "drum facility". In the stationary wheel method, a test tire is fixed on a stationary axle and rolls on a rotating drum. The test tire is installed in a way that it can roll against either the inside or the outside of drum (see Figure 2.22). In the stationary pavement method, test tire is usually mounted on a rotating arm and rolls around a stationary drum. The moulded replicas of road surfaces are fitted onto the drum in segments to simulate the interactions between tires and real pavements. Near- field measurements can be conducted with microphones installed in a configuration similar to the CPX or OBSI method.
59 Laboratory measurements do not need test vehicles or real pavement sections.
They can be used to predict the acoustical properties of new pavement technologies even before the test sections are constructed (Scofield, 2009). It is suitable when high precision is required, because most test parameters can be strictly controlled in the tests, such as wheel loading, rolling speed and test temperature. The measurement of tire vibrations using laser-Doppler vibrometers can be conducted in the stationary wheel method, since the test set-up makes it possible to focus the laser. Implemented properly, laboratory measurements can be used to predict the coast-by measurements (Sandberg and Ejsmont, 1984). Nevertheless, laboratory measurements also have several disadvantages. It is very difficult to reproduce the pavement surface on a drum, even with the same construction procedures. This may reduce measurement accuracy when the method is used to predict acoustical performance of to-be- constructed pavement. Background noise (such as the noise generated from drum power unit and the sound reflected by surrounding objects) is another concern in this approach. Moreover, the drum curvature caused by restricted device dimension may result in a significant difference from the flat road surface in practice.