Field measurement and evaluation of vibration in different areas of a metro depot

Feng Qingsong, Zhang Yunlai, Jiang Jian, Wang Ziyu, Lei Xiaoyan and Zhang Ling

1. Engineering Research Center of Railway Environment Vibration and Noise, Ministry of Education, East China Jiaotong University, Nanchang 330013, China

2. Institute of Sound and Vibration Research, University of Southampton, Highfield, Southampton, Hampshire SO17 1BJ, UK

Abstract: This paper presents a study of the characteristics of a railway vibration at three key sections containing different track structures in a metro depot. The results show that the vertical and horizontal vibration acceleration levels are proportional to train speed. The Z-weighted vertical acceleration levels obtained showed that the vibration source strengths at the ballast foot of the testing line and the throat area were very close. The vibration attenuation at the repair line was larger than that of the testing line. In the throat area, the peak frequency of vibration obtained at the ballast foot (2.5 m) could be shifted to a lower frequency band by using polyurethane sleepers instead of standard concrete sleepers. Polyurethane sleepers can help to reduce vertical vibration in a frequency band of 0-10 Hz. The vibration levels would satisfy the limits given in the ISO2631-2-2003 (2013) for any location more than 5 m away from the source at the testing line and 2.5 m away from the source at the repair line and throat area.

Keywords: field measurement; metro depot; peak frequency; vibration acceleration; environmental impacts

1 Introduction

Urban rail transit is becoming increasingly important in China, due to a fast-growing urban population, a trend that has placed enormous pressure on urban transport systems. Compared to many other public transport methods, rail travel is more reliable and cost effective. Currently, the urban railway system has been developed in more than 30 Chinese cities, with a total mileage of more than 6730.27 km (Hanet al., 2020), and it is already playing an important role in easing traffic pressure in these cities.

A railway depot, which is for parking, checking, maintaining and repairing trains, is often built underground in a city, taking up a large amount of land. To improve the land utilization rate and provide a high return for the investment, superstructures are usually built on top of the depot. Much attention has been paid to planning and designing domestic metro construction. However, due to the diverse types of rail structures used in the metro depot, plus the complexity of vibration sources involved, the characteristics of the vibration generated in the depot site still need further study. These vibrations can be transmitted through structures (such as columns and walls) to the building on top of the depot and cause potential problems for the safe use of these buildings, as well as health and safety hazards for their occupants.

Much research has been carried out to investigate various aspects of the vibration caused by urban railways (Connollyet al., 2015; Leiet al., 2003; Liet al., 2008; Liu and Lian, 2002; Liuet al., 2013; López-Mendozaet al., 2017; Sanayeiet al., 2014; Triepaischajonsak and Thompson, 2015; Vogiatziset al., 2018; Yaoet al., 2014; Zhanget al., 2017), such as train-induced ground vibration (Connollyet al., 2015; Zhanget al., 2017), building vibration (López-Mendozaet al., 2017; Vogiatziset al., 2018) and soil-structure interaction (Sanayeiet al., 2014; Yaoet al., 2014). Zouet al. (2015, 2017 and 2018) conducted a large number of vibration measurements in buildings and on the ground surrounding a depot in Guangzhou. Results showed that the ground vibration attenuation was faster at frequencies higher than 50 Hz. The vibration measured in the buildings was mainly in the range of 20–50 Hz, reflecting the buildings′natural frequencies of vertical vibration. A numerical model, which took the foundation into consideration, was also established to predict the vibration level in the upper floors. Many scholars (He and Xie, 2016; Xieet al., 2013, 2016) studied vibrations in the superstructure above the depot. A refined finite element model was established based on field measurements, in order to study the transmission of vibration in the building and evaluate the vibration comfort level.

Empirical methods are often used in practice due to the difficulties of obtaining reliable estimates of ground properties and other parameters. A few empirical approaches have been used for investigating railwayinduced ground vibration (Madshuset al., 1996; Nelson and Saurenman, 1987). In the US, the Federal Transit Administration (FTA) guidance manual (Hansonet al., 2006) also recommends the use of an empirical method to predict ground-borne vibration associated with a transportation project.

Unlike many other sections of a railway system, a depot must handle a diversity of trains. In addition, the depot is normally divided into adjacent areas for different purposes. Geological conditions, track structures and train speed in these areas can be vary, too. All of these factors make the vibration from a depot much more complicated than vibration from other sections along a railway system.

Most of the previous studies focused on the overall levels of ground vibrations inside a depot, and little attention has been paid to actual characteristics of the vibrations (which can be influenced by variation in track structure). The vibration transmission and environmental impact were often ignored as well. Based on field measurement data, this paper investigated the characteristics of vibration measured under various working conditions for different sections of a depot. The transmission of vibration in the depot and the resulting environment impact were subsequently studied. The results provided valuable knowledge for applying further vibration control in that particular depot.

2 Overview of the metro depot and arrangement of measuring locations

Field measurements were conducted in the train testing area inside the Guangzhou metro depot. The main structures/sections inside this area include the repair line, the testing line, the adjustment and engineering garage, the material warehouse and the operation complex building. The testing line is mainly used for testing and examining trains. It is located on the east side of the depot, running from north to south just outside the repairing workshop (see Fig. 1). Covered ladder sleepers are used on the track along the testing line. The normal train speed in this section is 40 km/h. The testing line is connected to the throat area, which is a junction area in which trains are diverted from the entrance into different sections inside the depot, such as the repair shop and parking garage (see Fig. 2). There are many turnouts in this area; consequently, trains pass more frequently than is the case with other sections inside the depot. Shock pads were also added to the track in the throat area to reduce vibrations. Train speed here is about 15 km/h. The repair line exists for maintaining and repairing trains inside a workshop. The speed limit for trains inside the workshop is 5 km/h.

Measurements were carried out for two test sections, as seen in Fig. 3. Section 1 is chosen to cover one testing line outside the repair workshop and another testing line inside the repair workshop. In order to measure the vibration of these two lines, a large number of measuring points were set up. Section 2 is inside the throat area. Two types of sleepers, concrete and polyurethane, are used in this area. Therefore, two separate measurements were conducted to cover both cases. More information about the train line is shown in Table 1. The trains parked in this depot are all type B, with 4 carriages, and all are empty when entering and leaving the depot. See Table 2 for basic train parameters. Soil layer information provided by survey data taken for the vehicle depot area is shown in Table 3.

3 Instrument and signal processing

The data acquisition system used in the test was DATaRec 4 DIC24 from HEAD, as shown in Fig. 4. It works with ArtemiS which is a data processing software that can carry out the data collection and some post-processing procedures. One microphone and 4 accelerators were used. Details of these transducers can be found in Table 4 and Table 5.

Table 6 ISO Suggested vibration limits in buildings

Table 7 Z vibration levels of source loads for four cases

Fig. 5 Point on the rail near the testing line (Location A1)

Fig. 6 Point on the rail near the throat area (Location B1)

Fig. 7 Point on the rail near the repair line (Location D1)

Fig. 8 Point on the sleeper near the testing line (Point A2)

Fig. 9 Point on the sleeper near the throat area (Point B2)

Fig. 10 Point on the sleeper near the repair line (Point D2)

Fig. 11 Near field in the testing line, 2.5 m

Fig. 12 Far field in the testing line, 17 m

Fig. 13 Acceleration level on the ground near the testing line

Fig. 14 Acceleration in the vertical direction in throat area (Case 1)

Fig. 15 Acceleration in the horizontal direction in the throat area (Case 1)

Fig. 16 Acceleration level on the ground near the throat area (Case 1)

Fig. 17 Acceleration in the vertical direction in the throat area (Case 2)

Fig. 18 Acceleration in the horizontal direction in the throat area (Case 2)

Fig. 19 Acceleration level on the ground near the throat area (Case 2)

Fig. 20 Near the repair line, 10.5 m

Fig. 21 Near the repair line, 19.5 m

Fig. 22 Acceleration level on the ground near the repair line

Table 4 Main technical specifications of PCB

Table 5 Main technical specifications of type 941B

Fig. 4 Instrument (a) and (b)

Table 1 Basic information regarding the train line in the testing line of the metro depot

Table 2 Calculation parameters of a metro type B vehicle

Table 3 Parameters of the soil layer in the Guangzhou metro depot

Fig. 1 Rail types in metro depot

Fig. 2 Plan view

Fig. 3 The layout of measurement locations

Some signal processing procedures were carried out before any further data analysis. These included (1) filtering and de-noising, (2) eliminating trend items, (3) appling Hanning and linear average to the data.

The verticalZ-vibration level is defined as (ISO 2631-2-2013, 2013):

4 The analysis of the track structure

4.1 Rail vibration

Figures 5–7 show the time history and frequency spectrum of vertical and horizontal accelerations measured on the rail at the testing line, the throat area and the repair line. From the figures:

(1) In the time domain, on the testing line, the maximum vertical acceleration of the rail is 129 m/s2, and the maximum horizontal acceleration is 93 m/s2. They are 43.1 m/s2and 30.7 m/s2, respectively in the throat area, 12.5 m/s2and 8.5 m/s2, respectively, in the repair line. The ratio of acceleration peaks at these tree sites is approximately 10.5 : 3.5 : 1. The difference is due to the fact that the acceleration value of the rail vibration is proportional to the speed of the train. The testing line has the highest train speed and the repair line has the lowest among these three sites.

(2) In frequency domain, for the testing line the vertical vibration shows multiple peaks between 10 Hz and 200 Hz, and has high amplitude in the range from 500 Hz to 1500 Hz. Compared to the vertical vibration, the amplitude of the horizontal vibration is much lower, with the highest amplitude around 300 Hz.

(3) The frequency spectrum of vertical and horizontal vibrations measured at the throat area are similar, with the horizontal vibration having a slightly smaller amplitude. They both have multiple peaks between 10 Hz and 200 Hz, and the main energy is between 700 Hz and 1500 Hz.

(4) On the repair line, the vertical vibration shows multiple peaks between 30 Hz and 1000 Hz, with the highest peak at 450 Hz. However, the maximum value of the horizontal vibration amplitude appears at 150 Hz. Due to the attenuation, the amplitudes of the vibration in both directions are reduced significantly after 1000 Hz.

4.2 Sleeper vibration

Figures 8–10 show the time history and frequency spectrum of vertical and horizontal accelerations measured on the sleeper at the testing line, the throat area and the repair line. From the figures:

(1) At the testing line, the maximum vertical acceleration measured on the sleeper is 10.73 m/s2, and the maximum horizontal acceleration is 11.7 m/s2. They are 2.45 m/s2and 12.35 m/s2, respectively, on the sleeper in the throat area, and 0.93 m/s2and 4.14 m/s2, respectively, on the sleeper from the repair line. For all cases, the vertical acceleration values of the sleeper are smaller than those of the horizontal direction.

(2) Although the acceleration values measured on the sleepers are positively correlated to the speed of train, the frequency spectrums of acceleration in the horizontal direction in these three cases are very different. This is mainly due to the fact that the structure and material of the sleeper used at these locations are different. The ladder sleeper is used on the testing line and the concrete sleeper is used in the throat area. An overhead concrete column structure is used on the repair line.

(3) In frequency domain, at the testing line the vertical acceleration of the ladder sleeper is relatively high–between 40 Hz and 150 Hz–and the horizontal acceleration has a high amplitude: between 40 Hz and 100 Hz, as well as between 200 Hz and 300 Hz.

(4) In the throat area, the vertical acceleration of the standard concrete sleeper has a high amplitude (below 200 Hz) and the horizontal acceleration shows a very high peak, of 100–200 Hz.

(5) The maximum horizontal vibration acceleration on the sleeper in the throat area is greater than that of the sleeper on the testing line. This is because the passing train excited the resonance of the angle steel bracket, which has the horizontal direction sensor mounted on it before it is glued to the sleeper by using glass glue. (GB 50868, 2013; Li, 2012; ISO 2631-2-2003, 2003).

(6) On the repair line, the main energy of vertical acceleration is below 200 Hz. The amplitude of the horizontal acceleration is significantly larger than that of the vertical direction, with a major peak of 120 Hz.

5 The analysis of ground measurements

5.1 Ground-borne vibration near the testing line

Figure 11 shows the time-history and frequency spectrum of the vertical and horizontal acceleration on the ground measured at the foot clamping (Location A3) of the testing line. As seen from the figure, the vertical and horizontal accelerations are very similar. In the time domain, both of the peak values are around 2 m/s2. In the frequency domain, the dominant frequencies of the vertical vibration are between 50 Hz and 100 Hz, with the highest value appearing at 76 Hz. The dominant frequencies of the horizontal vibration are between 40 Hz and 100 Hz, with the highest value appearing at 68 Hz.

Figure 12 shows the time-history and frequency spectrum of the vertical and horizontal acceleration on the ground, measured at a location farther away from the track (Location A5). In the time domain, the amplitude of vertical acceleration is larger than that for the horizontal direction. The maximum vertical acceleration is about 0.037 m/s2, and the maximum horizontal acceleration is about 0.021 m/s2. In the frequency domain, the dominant frequencies of the vertical vibration are between 20 Hz and 100 Hz, with the maximum peak appearing at 78 Hz and two other large peaks at around 40 Hz and 63 Hz. The dominant frequencies of the horizontal vibration are between 20 Hz and 90 Hz, with the maximum peak appearing at 23 Hz and two other large peaks at around 40 Hz and 63 Hz.

In comparing Figs. 11 and 12, high-frequency components are attenuated faster than low-frequency components, and vibration in the horizontal direction is attenuated faster than that of the vertical direction with distance. In the far field, an amplitude in a 1/3 octave band of 50 Hz obviously decreases. This is in alignment with Zouet al. (2015).

Figure 13 shows the vertical and horizontal acceleration on the ground measured at different distances from the track in one-third-octave bands. As the figure shows, the vibration level decreases with distance. The maximum level of vertical acceleration is reduced from 96 dB at 2.5 m to 55 dB at 25 m. The maximum level of horizontal acceleration is reduced from 96 dB at 2.5 m to 50 dB at 25 m. The reduction at higher frequencies (> 80 Hz) is much more significant compared to the lower frequencies.

5.2 Ground-borne vibration in the throat area (Case 1)

Figures 14–15 show the time-history and frequency spectrum of the vertical and horizontal acceleration on the ground, measured at two locations, 2.5 m (B3) and 11.5 m (B5) away from the track in the throat area (Case 1). From the figures:

(1) In the time domain, the peaks of acceleration in the vertical and horizontal directions are about 0.78 m/s2and 0.32 m/s2, respectively, at 2.5 m, and about 0.038 m/s2and 0.051 m/s2, respectively, at 11.5 m.

(2) In the frequency domain, at 2.5 m, the dominant frequencies of the vertical vibration are between 20 Hz and 80 Hz, with the highest peak appearing at 60 Hz and two other very high peaks at 33 Hz and 71 Hz. The dominant frequencies of the horizontal vibration are between 20 Hz and 100 Hz, with the highest value appearing at 62 Hz and another very high peak at 33 Hz. At 11.5 m, the dominant frequencies of the vertical vibration are between 20 Hz and 80 Hz, with the highest peak appearing at 33 Hz and two other very high peaks at 43 Hz and 62 Hz. The dominant frequencies of the horizontal vibration are between 20 Hz and 80 Hz, with the highest value appearing at 32 Hz and two other very high peaks at 25 Hz and 49 Hz. Comparing the frequency spectrum at 2.5 m and 11.5 m, the first peak frequency is 0–10 Hz, but at 2.5 m, the first peak frequency band (0–10 Hz) of vertical vibration is obviously higher than for other case conditions.

Figure 16 shows the vertical and horizontal acceleration on the ground, measured at different distances from the track in one-third-octave bands. From the figures, it can be seen that the maximum level of vertical acceleration was reduced from 88 dB at 2.5 m to 65 dB at 11.5 m. The maximum level of horizontal acceleration changed from 85 dB at 2.5 m to 70 dB at 7.5 m, but then was reduced to 59 dB at 11.5 m.

5.3 Ground-borne vibration in the throat area (Case 2)

Figures 17-18 show the time-history and frequency spectrum of the vertical and horizontal acceleration on the ground, measured at two locations, 2.5 m (C3) and 11.5 m (C5) away from the track in the throat area (Case 2). From the figures, it can be see that in the time domain, the peaks of acceleration in the vertical and horizontal directions are about 0.95 m/s2and 0.297 m/s2, respectively, at 2.5 m, and about 0.092 m/s2and 0.085 m/s2, respectively, at 11.5 m.

In the frequency domain, at 2.5 m, the dominant frequencies of the vertical vibration are between 20 Hz and 80 Hz, with the highest peak appearing at 50 Hz, and two other very high peaks at 33 Hz and 62 Hz. The dominant frequencies of the horizontal vibration are between 20 Hz and 80 Hz, with the highest value appearing at 48 Hz and another very high peak at 27 Hz.

At 11.5 m, the dominant frequencies of the vertical vibration are between 20 Hz and 80 Hz, with the highest peak appearing at 45 Hz and two other very high peaks at 32 Hz and 75 Hz. The dominant frequencies of the horizontal vibration are between 20 Hz and 80 Hz, with the highest value appearing at 29 Hz and another very high peak at 35 Hz.

From Figs. 17–18, it is shown that the acceleration in the horizontal direction is lower than that of vertical direction. High-frequency components attenuate much faster than low-frequency components. As a result, the main peaks in the frequency domain move toward a lowfrequency direction when the measurement location is farther away from the track.

Figure 19 shows the vertical and horizontal acceleration on the ground, measured at different distances from the track in one-third-octave bands. For most of the frequency bands, the acceleration level decreases with distance. The maximum level of vertical acceleration is reduced from 90 dB at 2.5 m to 70 dB at 21.5 m. The maximum level of horizontal acceleration changes from 78 dB at 2.5 m to 67 dB at 11.5 m, but then increases to 80 dB again at 21.5 m. Compared to the test point in the throat area (Case 1), at 11.5 m, the vertical and horizontal vibration levels are higher, and the corresponding frequency of main peak value is higher. This indicates that when polyurethane sleepers are used the vibration at 11.5 m is amplified.

Compared with the throat measuring point in Case 1, for which concrete sleepers are used, the vertical and horizontal vibration acceleration for Case 1, at 2.5 m, are very close to the values of Case 2, for which polyurethane sleepers are used. The peak values of vertical and horizontal vibration acceleration for Case 1, at 11.5 m, are smaller than those for Case 2. This can be due to the quality of polyurethane sleepers being lower than that of concrete sleepers, which results in the vibration of polyurethane sleepers being greater than that of concrete sleepers at farther distances from the track.

In the throat area, the main frequency band of vibration energy decreases with an increase in the distance from the line, and the vibration attenuation is obvious with an increase in distance from the near point of the line. The vibration attenuation is not obvious when the distance from the line is farther, especially for those less than 60 Hz. Vibration attenuation is very small as distance increases.

5.4 Ground-borne vibration near the repair line

Figure 20 shows the time-history and frequency spectrum of vertical and horizontal acceleration on the ground, measured at Location D4, inside the repair workshop. In the time domain, the maximum values of vertical and horizontal acceleration are about 0.025 m/s2and 0.016 m/s2, respectively. In the frequency domain, the dominant frequencies of vertical vibration are between 20 Hz and 80 Hz, with the highest value appearing at 32 Hz and another major peak at 71 Hz. The dominant frequencies of the horizontal vibration are between 20 Hz and 80 Hz, with the highest value appearing at 62 Hz and three other peaks at 21 Hz, 32 Hz and 39 Hz.

Figure 21 shows the time-history and frequency spectrum of the vertical and horizontal acceleration on the ground, measured at Location D5 outside the repair workshop and farther away from the track. In the time domain, the maximum values of vertical and horizontal accelerations are about 0.0035m/s2and 0.0037m/s2, respectively. In the frequency domain, the dominant frequencies of vertical vibration are between 20 Hz and 80 Hz, with the highest value appearing at 65 Hz. The dominant frequencies of horizontal vibration are between 20 Hz and 80 Hz, with the highest value appearing at 24 Hz and three other peaks at 40 Hz, 53 Hz and 61 Hz.

From Figs. 20–21 it can be seen that the attenuation of the high frequency is more obvious as the distance from the track increases. It is worth noting that horizontal acceleration is lower than is the case with the vertical acceleration.

Figure 22 shows the vertical and horizontal acceleration on the ground measured at different distances from the track in one-third-octave bands. From the figure it can be seen that the vibration level decreases with distance. The maximum level of vertical acceleration is reduced from 71 dB at 2.5 m to 53 dB at 19.5 m. The maximum level of horizontal acceleration is reduced from 70 dB at 2.5 m to 46 dB at 19.5 m. In most of the frequency bands, the vibration level decreases with distance, except for the near filed measurement pint (Point D3). Comparing to the near filed measurement pint, the high frequency vibration of the other two far field measuring points are attenuated more significantly, beyond 40 Hz.

6 The evaluation of the ground environment

ISO2631-2-2003 (2003) gives the limit of building vibration using vibration intensity as well as human exposure time, as shown in Table 6.

Table 7 gives theZ-weighted vibration levels at the ballast foot and on the ground for four cases. From Table 7, theZ-level at the ballast foot position is the largest at the test line, followed by two others in the throat area and the smallest for the repair line. Generally speaking, theZ-level at the ballast foot is positively correlated with train speed. ForZ-levels on the ground, the largest value is observed in the throat area, followed by the test line and the repair line. This is because the track in the throat area was built on a concrete foundation, with smaller attenuation, while the test line was built directly on soil, with larger attenuation. The overhaul line is indoors, with the smallest running speed, and hence, the smallestZlevel.

Tables 8–11 and Fig. 23 show theZ-weighted vertical acceleration levels from five data sets measured at the three testing locations. The gray area in the figure indicates the variance of the data used in the analysis. The red curves show the average values. The blue dashed line is the linear fitting curve of vibration level attenuation with distance.

Fig. 23 The degree of Z direction vibration under different cases

Fig. 24 The limit of the Z direction vibration under working conditions

Table 8 Vertical Z vibration levels at all points of the test line (dB)

Table 9 Vertical Z vibration levels at all points of the throat area (Case 1) (dB)

Table 10 Vertical Z vibration levels at all points of the throat area (Case 2) (dB)

Table 11 Vertical Z vibration levels at all points of the repair line (dB)

Table 12 gives the parameters used to calculate the linear fitting of the five data sets, whereYis the acceleration level (dB) andXis the distance from the vibration source (m).

As listed in Table 10, the vibration source strength obtained for the testing line (79.154 dB) is close to that of the throat area (78.492 dB and 79.175 dB). The vibration source strength for the repair line, 69.823 dB, is the lowest for all cases. Regarding the attenuation rate, that of the vertical direction vibration is 1.262 dB/m for the testing line and 1.418 dB/m for the repair line. The attenuation rate for the repair line is 0.15 dB larger than that of the testing line, due to a faster train speed on the testing line (40 km/h) than is he case on the repair line (5 km/h). In the throat area, the two lines are basically the same except for the sleeper materials. The attenuation rate of ground vibration for Case 1 (the concrete sleeper) (1.460 dB/m) is much larger than that for Case 2 (the polyurethane sleeper) (0.975 dB/m). The results indicate that the polyurethane sleepers can cause a larger vibration with less attenuation.

Table 12 Z-weighted vertical acceleration vibration

Figure 24 compares the vibration limits (ISO2631-2-2003) with measurements for four working conditions. Treating the depot as a residential area, the daytime limit given in the standard is 80–86 dB (80 dB is used here) and the night limit is 77 dB. The vibration level can exceed the limits within a radius of 5 m for the testing line and 2.5 m for both the repair line and throat area. If any vibration within the radius causes concerns, a vibration damping pad for track and vibration isolation groove structure are recommended for the throat area and the testing line, and the vibration reduction fastener is recommended for the repair line.

7 Conclusions

Based on the field measurement carried out at a metro depot in Guangzhou, China, this paper analyzed the characteristics of rail and ground vibration as measured in three areas with different track structures, train speeds, etc. A subsequent environmental impact assessment was conducted for all cases. From the results:

(1) The characteristics of vertical and the horizontal vibrations obtained from the testing line, the throat area and the repair line are very different. The acceleration value of rail vibration is proportional to train speed. The larger the speed, the larger the rail acceleration level. Therefore, the largest peak rail acceleration value was observed on the testing line, which had highest train speed. The main frequency range of vertical vibration for the testing line and the throat area were 300–2000 Hz. The main frequency range of horizontal vibration for the testing line was 100–1000 Hz, and 300–2000 Hz for the throat area. On the repair line, the main frequency range of both vertical and horizontal vibrations was 40–300 Hz.

(2) The intensities of vibration source caused by a train on the test line and the throat area were very close, and that of repair line was the lowest. In the throat area, when polyurethane sleepers were used, because the quality of polyurethane sleepers is lower than that of concrete sleepers, the vibration of sleepers leaving the track was greater than that of the concrete sleepers, and the decay rate was slower. Therefore, polyurethane sleepers are not conducive to vibration reduction, and vibration reduction measures need to be further optimized.

(3) Compared to the limits given by ISO2631-2-2003, the vibration level on the test line exceeded the limit in a radius of 5 m from the vibration source. The distance was 2.5 m for the throat area and the repair line.

Acknowledgement

The authors wish to express their gratitude to the National Natural Science Foundation of China (Grant Nos. 52068029, 51878277, 52178423), the Major Discipline Academic and Technical Leaders Training Program of Jiangxi Province Youth (Grant No. 20194BCJ22008), the Key Research and Development Program of Jiangxi Province (Grant No. 20192BBE50008) for supporting this work.