Measurement Discrepancies in OBSI Testingunder Varying Testing Speeds and Pavement Sections

Abstract

The on-board sound intensity (OBSI) method has become a standard tool for evaluating tire–pavement interaction noise. However, the common assumption that the difference of sound intensity level (SIL) measured by front and rear microphones ( Δ SIL ) can be averaged or neglected affects the comparability of results across studies. In this study, OBSI noise measurement was conducted on two types of asphalt pavements, asphalt concrete with three different nominal maximum aggregate sizes and stone mastic asphalt, at speeds of 30, 40, 50, and 60    km / h . The overall SIL, the 1/3-octave-band spectra within frequency range from 500 to 5,000 Hz, and specified frequency ranges with 1,600 Hz as a critical value (i.e., the mid-low-frequency range from 500 to 1,600 Hz and the high-frequency range from 1,600 to 5,000 Hz) were recorded. Based on this, the measurement differences between the front and rear probes in OBSI testing method were quantified and analyzed. The results show that SIL results from front probes generally were higher than those from rear probes, with numerical differences ranging from 1.3 to 2.7 dBA. The overall SIL exhibited a positive linear relationship with the logarithm of testing speeds, and linear fitting by results from rear probes indicated lower intercepts but steeper slopes than front probes, and Δ SIL had a negative liner correlation, achieving a minimum at 60    km / h . The 1/3-octave-band spectra of Δ SIL indicated that a common minimum occurred at about 1,600 Hz regardless of pavement types, nominal maximum aggregate sizes, and testing speeds. This implies that 1,600 Hz may be a critical frequency of vibration-induced noise and aerodynamic noise, thus validating the rationality of regarding the 1,600 Hz as the critical frequency. This study emphasizes that using only the result from a single probe or a simple average based on the front and rear probes may introduce deviations, thus inducing different conclusions. Improving the standard speed also is a way to reduce testing deviations for OBSI testing method.