Study on Influencing Factors and Spectrum Characteristics of Tire/Road Noise of RIOHTrack Full-Scale Test Road Based on CPXT Method

Abstract

In order to investigate the influence of different tire textures, pavement types, and vehicle parameters on the tire/road noise level and its spectrum characteristics, 19 kinds of asphalt pavement main structures of RIOHTrack full-scale test track were tested by the close-proximity trailer (CPXT) tire/road noise detection method. Considering investigated parameters such as tire texture, vehicle speed, and trailer axle weight, and relying on multi-functional road condition rapid detection vehicle and laboratory tests to collect a variety of road surface information and material parameters, a multiple-linear-regression model of tire/road surface noise level of RIOHTrack (Research Institute of Highway Full-scale Test Track) asphalt pavement was constructed. Finally, the causes of noise level differences among different influencing factors were further analyzed through spectrum characteristics. The results show that vehicle speed is the most important factor affecting tire/road noise. The noise level of different tires varies due to different textures, but the noise level among different trailer axle weights is roughly the same. Vehicle speed (v), FWD center deflection (D₀), surface asphalt mixture air voids (VV), sensor-measured texture depth (SMTD) and international roughness index (IRI) were selected to establish the noise prediction models of different tire textures. Noise spectrum analysis shows that the spectrum of different vehicle speeds is significantly wide in the full frequency range, and the spectrum variation of differently textured tires is mainly concentrated in a certain range of the peak frequency. The noise spectrum curve of porous asphalt concrete (PAC13) is significantly lower than that of other asphalt mixtures in the full frequency range above 800Hz, indicating a greater noise reduction effect.

1. Introduction

In the context of the evolving transportation landscape, the requirements for road infrastructure have undergone a significant transformation. According to the stringent demands of the fifth-generation roads concept, roads are no longer merely expected to fulfill the fundamental criteria of “comfort, durability, and safety”. Instead, they are now required to embrace new and far-reaching connotations. These novel attributes include adaptability, which implies the ability of roads to adjust to various traffic conditions, vehicle types, and environmental changes; automation, enabling seamless integration with autonomous driving technologies; and weather resistance, ensuring consistent performance regardless of extreme weather events such as heavy rain, snowstorms, or intense heat. Among these advanced requirements, reducing driving noise stands as one of the most tangible and crucial manifestations [1].

Currently, traffic noise has emerged as a dominant form of environmental pollution within the transportation domain. It pervades the areas adjacent to roadways, seriously encroaching upon and degrading the quality of life of the residents living in these vicinities. Traffic noise can be classified into three main categories: tire/road coupling noise, vehicle power system noise, and aerodynamic noise. Extensive research has demonstrated that when the vehicle speed exceeds 40 km/h, tire/road coupling noise takes precedence as the primary contributor to traffic noise [2]. Donavan et al. found that 78% of the noise comes from tire–pavement interaction [3]. This finding has thrust the issue of how to effectively reduce tire/road coupling noise into the spotlight, making it a central and highly challenging topic that researchers both at home and abroad are earnestly endeavoring to address. They are exploring a multitude of innovative solutions, ranging from developing new tire materials and tread patterns to optimizing road surface textures, all in a bid to mitigate this persistent and troublesome source of noise pollution.

At present, the tire/road noise detection methods are mainly divided into the pass-through method and the close-proximity detection method. The pass-through method is further divided into the controlled pass-through method (CPB) and the statistical pass-through method (SPB). The controlled pass-through method uses a specified standard vehicle/tire combination to test the noise peak of the vehicle passing through, hence its name. The SPB statistical pass-through method is similar to the controlled pass-through method, but the traffic flow is different. The CPB method uses a specified standard vehicle/tire combination, while the SPB stipulates that the speed is not less than 50 km/h in the free traffic flow, from which no less than 180 vehicles are selected for testing. The tire/road noise measured by the pass-through method is not only the simple friction noise between the tire and the road, but also includes the propagation state of the noise in the air, so it is closer to the noise level of the entire environment. The close-proximity detection method uses a microphone installed close to the ground to measure the tire/road sound pressure. The sound pressure measures the sound power radiated by the sound source. The sound pressure can be tested by the difference in air density caused by the sound source and is measured using a standard sound level meter. The close-proximity method basically shields against the interference of external noise during the noise measurement process. Its measurement results are closest to the friction noise between the tire and the road surface, which is more conducive to the study of noise reduction technology for roads or tires.

In the late 1970s and early 1980s, various countries began to study asphalt pavement noise reduction technology and proposed three pavement noise reduction mechanisms: porous sound absorption, acoustic interference, and improving pavement elasticity. However, the mechanism of tire/road noise generation is relatively complex, and the influencing factors are also diverse, involving both the performance of tires and pavements and the interaction between tires and pavements [4]. Wang Xudong [5] compared six different noise-reducing pavement structures from the perspective of pavement structure and believed that particle size and void ratio are the main factors affecting pavement noise. Elisabete Freitas [6] established a tire–pavement noise model using data mining technology and believed that the most important parameters include vehicle speed, temperature, aggregate particle size, average texture depth, etc. Elisabete Freitas et al. [7] proposed a noise prediction model based on traffic flow speed, pavement type, and existing road surface defects. De Chen et al. [8] established a tire/road noise model at different vehicle speeds based on the surface characteristics and sound absorption coefficient of porous asphalt concrete. Dae Seung Cho et al. [9] evaluated the sound power levels according to surface type, vehicle type, and vehicle speed. Most studies mainly analyze the noise level under different factors from a macroscopic perspective, and few conduct in-depth research on the noise spectrum distribution under different conditions. In addition, domestic research on the trailer method for testing noise is not sufficient. Based on the RIOHTrack full-scale test road, this paper uses the trailer-type on-site close-proximity tire/pavement noise detection method (CPXT method) to test and study the noise and spectral characteristics of 19 asphalt pavement main structures, considering multiple factors.

2. Overview of Test

2.1. Pavement Structure Type

The road section for noise detection and analysis in this study is from the RIOHTrack full-scale test road [10]. RIOHTrack is China’s first accelerated loading test track, which conducts simultaneous and spatial comparative studies on the service performance of different pavement structures throughout their life cycle. The RIOHTrack is located at the highway traffic test site in Tongzhou District, Beijing, and it has a runway-type layout and a total length of 2039 m. There are 38 types of pavement structures in total, which provides reliable platform support for studying the tire/road noise characteristics of different pavement types in space and time. In order to avoid interference with the pavement noise test results during vehicle turns, this study only selected the test results of 19 main structures (STR1–STR19, all asphalt pavements) of the straight section as the data source for analyzing the impact of different variables on pavement noise (see Figure 1).

In terms of the type of surface layer materials, the 19 main structures used four different asphalt concretes, namely: SAC13-65, SAC13-70, SMA13, and PAC13. The surface layers were all made of basalt, and their maximum nominal particle size was 13.2 mm. STR9 was an open-graded porous asphalt concrete using high-viscosity asphalt named SBS HV4. Related studies have shown that it has a significant noise reduction effect [11]; the rest were dense asphalt concretes using SBS I-D modified asphalt, but the mineral aggregate grading was different. The specific material properties are shown in

Table 1. Asphalt pavement surface material types.

Mix Type	Structure Number (STR)	Coarse Aggregate Content (>4.75 mm, %)	Binder Type	Air Void(%)
SAC13-65	1, 2, 3, 10, 11, 12, 13, 14, 15	65	SBS I-D	4.6
SAC13-70	4, 5, 6, 7, 8	70	SBS I-D	5.1
SMA13	16, 17, 18, 19	75	SBS I-D	4.5
PAC13	9	80	SBS HV4	18.3

.

2.2. Methods

The CPXT (close-proximity trailer test) method is a method that can be widely used to evaluate road noise during vehicle driving [12,13,14]. It places a standard wheel (standard tire pattern) in a small trailer with sound insulation, arranges multiple microphones around the standard wheel, and measures tire/road noise during driving from different angles, as shown in Figure 2.

This test uses the counterweight tire/road noise trailer detection equipment independently developed by the Highway Research Institute of the Ministry of Transport [15]. The test tire can be replaced or directly removed before testing; the test wheel is connected to the trailer frame through a hinge link, and the load of the test wheel is adjusted by adjusting the counterweight at the end of the hinge link, thereby realizing the detection of tire/road noise under different load conditions. The mechanism of road noise generation is complex and diverse, and there are many factors that affect tire/road noise. After comprehensive consideration, three key variables, namely vehicle speed, tire texture, and trailer axle weight, were finally determined to test road noise. The variable parameters are shown in

Table 2. Summary of basic variable parameters of road noise.

Test Variables	Parameters
Speed	40 km/h, 60 km/h, 80 km/h
Tire texture	1#, 2#, 3#
Trailer weight	300 kg, 400 kg, 500 kg

.

2.2.1. Speed

The National Center for Asphalt Technology (NCAT) used the close-proximity (CPX) method to measure noise, and the results showed that for every km/h increase in speed on hot mix asphalt (HMA) pavement, tire-pavement noise increases by an average of 0.18 dBA per mile increase of vehicle speed for both hot mix asphalt (HMA) [16]. As mentioned above, when the vehicle speed exceeds 40 km/h, tire/road noise becomes the main noise source. Limited by the radius of curvature of the circular curve and safety factors, this study finally selected 40 km/h, 60 km/h, and 80 km/h for testing. Each speed was tested for 3 consecutive laps, and the average value was taken for analysis to ensure the reliability of test results.

2.2.2. Tire Texture

The tread shape of the tire is also an important factor affecting tire/road noise. In the selection of standard tires, through the investigation of the current tire market, tires that have patterns as close as possible to international standard tires and are widely used are chosen as reference tires for noisy vehicles. The texture of the tires is shown in Figure 3.

2.2.3. Trailer Axle Weight

The weight of the vehicle may also be a potential factor affecting the noise level. A special counterweight is added to the trailer to increase the axle weight of the trailer. After counterweighting, a tire weighing instrument is used to calibrate the axle weight of the test wheel to 300 kg, 400 kg, and 500 kg, respectively (Figure 4).

3. Results

3.1. Comparison of Sound Pressure Level (SPL)

3.1.1. Effect of Vehicle Speed

As can be seen from Figure 5, different surface materials and tire textures have obvious differences at different vehicle speeds. As the vehicle speed increases, the noise level increases to varying degrees, but the increase is not linear and tends to weaken significantly at 80 km/h. When the vehicle speed doubles from 40 km/h to 80 km/h, the noise level increases by about 10dB (A). The results show that the vehicle speed has a significant impact on the noise level.

As the vehicle accelerates, the kinetic energy generated at the interface between the tire and the road surface escalates exponentially. The increased rotational speed of the tires leads to more rapid and forceful interactions with the irregularities of the road texture, creating a higher-intensity vibration that translates into a significantly louder noise output.

3.1.2. Effect of Tire Texture

From the test results, we can see that different tire textures have different effects on noise levels (Figure 6). The noise level is ranked as 1# > 2# > 3#. The use of 3# tires can produce a lower noise level on roads with different surface layers. The tire tread area mainly affects the tire vibration noise, and the tire tread depth, width, and spacing and number of grooves mainly affect the pumping noise. Therefore, among the three types of tires, this type of pattern of 3# tires can effectively reduce the comprehensive noise level of tires/road surfaces, and is recommended from the perspective of noise reduction.

Tires with different textures exhibit varying degrees of acoustic behavior. Some tire textures feature deeper grooves and more aggressive tread patterns, which can disrupt the airflow and create turbulence as the tire rotates. This turbulence, in turn, contributes to an elevated noise level. Conversely, tires with smoother textures or more optimized tread designs are better able to reduce the generation of such aerodynamic disturbances, resulting in a quieter operation. This discovery has opened up new avenues for manufacturers to design tires that not only provide excellent traction and durability but also offer enhanced acoustic comfort.

3.1.3. Effect of Trailer Axle Weight

The results show that the noise level is not sensitive to vehicles with different axle weights. From the sound pressure level of 300 kg to 500 kg, the noise level of the same surface layer material under different axle loads is roughly the same (see Figure 7). One possible reason is that within this axle weight range, the difference in the contact state between the tire and the road is very small.

While an increase in axle weight does result in a greater load being applied to the tires and the road, the effect on the noise-generating mechanisms is minimal. The fundamental factors governing tire/road noise, such as the interaction between the tire tread and the road surface texture and the aerodynamic forces acting on the tire, are not significantly altered by changes in axle weight within the normal range of vehicle operation. Even when heavier loads are considered, the noise levels do not show a substantial and consistent increase. Further analysis of its spectrum is needed to explore in depth the impact of axle load on road noise.

3.2. Construction of Noise Level Model

It is generally believed that there are two major factors that affect tire/road noise: vehicle factors and road factors. The literature shows that various influencing parameters can affect tire/road noise, and different noise estimation models can then be established based on the different factors [6,17,18,19,20,21,22,23,24,25]. This study selected three different variables for vehicle factors: different vehicle speeds, different tire textures, and different trailer axle weights. The road factors encompass the characteristics of the surface material and the basic information of the road surface. For the characteristics of the surface material, the void ratio of the surface-layer asphalt concrete is the main indicator due to the important influence of the void ratio on road noise. The basic information of the road surface of 19 main structures of RIOHTrack was tested using the road condition rapid detection system (called CICS) to obtain relevant indicators.

In view of the influence of the above-mentioned different variables on noise, the fitting method of multiple linear regression was used to firstly make the magnitude of each variable consistent, and then establish the functional relationship between the SPL of different tires and each variable (see

Table 3. Multiple-linear-regression model of noise level.

Tire No.	Multiple-Linear-Regression Model	R2
1#	$\mathrm{SPL}=32.090+0.080{\mathrm{D}}_0-0.972\mathrm{SMTD}+0.041\mathrm{IRI}-0.078\mathrm{VV}+35.052{\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{v}$	0.982
2#	$\mathrm{SPL}=23.709+0.023{\mathrm{D}}_0-0.772\mathrm{SMTD}+0.138\mathrm{IRI}-0.071\mathrm{VV}+39.395{\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{v}$	0.975
3#	$\mathrm{SPL}=26.944+0.046{\mathrm{D}}_0-0.829\mathrm{SMTD}+0.031\mathrm{IRI}-0.071\mathrm{VV}+37.068{\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{v}$	0.984

).

In the table, D₀ is the FWD center deflection (0.01 mm), SMTD is the standard mean structural depth (mm), IRI is the international roughness index (m/km), VV is the void ratio (%), v is the vehicle speed (km/h), and SPL is the sound pressure level.

From the regression model, we can see that the coefficient of determination of the model is very high (>0.97), and the coefficients of these indicators are similar in size and the ± signs are consistent, indicating that this model can reflect the influence of different indicators on the noise level to a certain extent. In the model, the road noise is negatively correlated with the road surface structure depth (SMTD) and the road surface material air void (VV), and is positively correlated with the vehicle speed (v), the international roughness index (IRI), and the center deflection (D₀). The vehicle speed is still the main factor determining the noise level. This model provides a certain reference for the technical path to effectively reduce the tire/road noise level.

After a comprehensive and in-depth test analysis of the experimental variables related to noise detection, as well as the characteristics of surface materials and the basic information of the road surface obtained from the measurements conducted by the road condition inspection vehicle, a significant research achievement was realized. Leveraging the unique and highly controlled environment of the RIOHTrack full-scale test road, a series of sophisticated noise prediction models were constructed.

These models were specifically tailored for three distinct types of tire textures, aiming to accurately forecast the tire/road noise under various road and vehicle operation conditions. The construction of these models was based on several crucial parameters. The FWD (falling weight deflectometer) center deflection (D₀), which reflects the load-bearing capacity and structural integrity of the pavement structure, plays a vital role. A greater D₀ value might indicate a more flexible road surface, which could interact differently with the tires and contribute to noise generation.

The standard mean structural depth (SMTD), an indicator of the road surface texture’s roughness in terms of its three-dimensional profile, is another key factor. A coarser SMTD implies that when sound waves are transmitted into the material, they will be continuously reflected and refracted in the pores. The energy of the sound waves is gradually consumed in this process and converted into heat energy, thereby playing a role in sound absorption and noise reduction.

The international roughness index (IRI), a widely recognized measure of road surface roughness that quantifies the vertical deviations of the road surface profile, provides an overall assessment of the road’s smoothness. A lower IRI value corresponds to a smoother road, generally resulting in less noise from tire–road interactions.

The surface material air void (VV) also contributes to the model. The air voids within the road surface material can affect the acoustic properties of the road, as they can either absorb or reflect sound waves generated by the tires.

Finally, vehicle speed (v), previously established as a dominant factor influencing tire/road noise, was incorporated into the models. As the vehicle speed increases, the dynamic forces at the tire–road interface change, and these changes need to be accounted for in the prediction models.

By integrating these multiple parameters, the constructed noise prediction models for the three different tire textures offer a more comprehensive and accurate way to understand and anticipate tire/road noise, which can serve as a valuable tool for road designers, tire manufacturers, and environmental researchers in their efforts to mitigate traffic-related noise pollution.

3.3. Noise Spectrum Analysis

From the perspective of the total sound pressure level of noise, only analyzing the absolute value of the noise level cannot entirely explain the noise difference, and noise spectrum analysis is an important supplement to the noise level evaluation. Some studies have investigated the spectral noise characteristics of the car/truck on pavement [26,27,28,29]. The spectrum analysis in this paper uses a 1/3 octave band, with a frequency range from 16 Hz to 20,000 Hz. In order to more intuitively perceive the sound pressure level characteristics at different frequencies, logarithmic coordinates are used for comparative analysis.

3.3.1. Spectrum Analysis at Different Vehicle Speeds

It can be seen from the noise spectrum that no matter which type of road surface, as the vehicle speed increases, the spectrum curve gradually moves upward (see Figure 8). In the low-frequency band, the spectrum curve is flat at a speed of 40 km/h, while a peak appears at 63 Hz at 60 km/h and 80 km/h, reflecting the difference between high speed and low speed; in the high-frequency band, the three spectrum curves are basically parallel, and the curves of 60 km/h and 80 km/h are closer.

3.3.2. Spectrum Analysis with Different Tires

For different tire textures, the difference is mainly reflected in the frequency range of 630~2000 Hz, which is also the peak zone of the spectrum curve (Figure 9). Different textures cause the peak and shape of this section to change. The spectrum of the 3# tire is the flattest, which is also the main reason for its relatively low noise level.

3.3.3. Spectrum Analysis for Different Trailer Axle Weights

From the noise spectrum analysis, the influence of different axle loads on the noise level is not significant (Figure 10). At three vehicle speeds, the noise spectra of different axle loads are basically overlapped. This result shows that within the axle load range of 300 kg~500 kg, the noise level is basically maintained in a stable state and is little affected by the change of axle load.

3.3.4. Spectrum Analysis for Different Surface Materials

There are four different surface materials in this test, of which the STR9 with porous asphalt concrete (PAC13) section presents the greatest noise reduction effect. From the spectrum distribution diagram in Figure 11, it can be seen that the spectrum curve of the PAC13 section is basically the same as that of the other three surface materials within the frequency range less than 800 Hz. However, the noise level is significantly lower than that of the other three asphalt mixtures in the full frequency range greater than 800 Hz, indicating that it has an effective noise reduction function.

4. Conclusions

Some conclusions based on our testing and analysis can be summarized as follows:

(1)

From the perspective of noise level, vehicle speed is the most important factor affecting tire/road noise; the noise levels of different tire textures also vary to a certain extent, and the tire/road noise level can be reduced by changing the tire texture; different axle weights have little effect on noise.

(2)

Through test analysis of experimental variables of noise detection, characteristics of surface materials, and basic information of road surface tested by road condition inspection vehicle, noise prediction models for three different tire textures were constructed based on the RIOHTrack full-scale test road using FWD center deflection (D₀), standard mean structural depth (SMTD), international roughness index (IRI), surface material air void (VV), and vehicle speed (v).

(3)

Analyzing noise spectra, the spectrum curves of different vehicle speeds are basically parallel, and the spectrum lines gradually move upward as the vehicle speed increases; the differences in different tire textures mainly exist in the spectrum peak range (630 Hz~2000 Hz), which in turn affects the noise level; and the spectra of different trailer axle weights basically overlap, which explains why the noise levels of different axle weights are roughly the same.

(4)

In the full frequency range greater than 800 Hz, the noise spectrum curve of porous asphalt concrete PAC13 is significantly lower than that of the other three asphalt concretes, which fully demonstrates its significant noise reduction effect.