- Email: [email protected]
Field study on high-speed train induced fluctuating pressure on a bridge noise barrier
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Contents lists available at ScienceDirect
Journal of Wind Engineering & Industrial Aerodynamics journal homepage: www.elsevier.com/locate/jweia
Field study on high-speed train induced fluctuating pressure on a bridge noise barrier Xiao-hui Xiong a, b, c, Ai-hua Li a, b, c, Xi-feng Liang a, b, c, Jie Zhang a, b, c, * a b c
Key Laboratory of Traffic Safety on Track of Ministry of Education, School of Traffic & Transportation Engineering, Central South University, Changsha 410075, China Joint International Research Laboratory of Key Technology for Rail Traffic Safety, Changsha, Hunan 410075, China National & Local Joint Engineering Research Center of Safety Technology for Rail Vehicle, Changsha, Hunan 410075, China
A R T I C L E I N F O
A B S T R A C T
Keywords: Fluctuating pressure Field measurement Noise barrier High-speed train Aerodynamics
The pressure variations induced by a CRH380A EMU on a 2.15 m high bridge noise barrier are investigated in a field measurement. The familiar fluctuating pressure time history curves and the peak-to-peak pressure (ΔP) distributions attributing to the train head on the barrier surfaces are presented. A comparison of the positive head pulse pressure is made between the measurement results and the data calculated by empirical equations in EN 14067-4. Furthermore, the influences of train speed, train running lines, locations of measurement points, train marshalling length and environmental wind speed on ΔP are analysed. The results indicate as the train speed increases, the corresponding time intervals of ΔP decrease gradually, whereas their slopes become increasingly steep. For a CRH380A EMU, the aerodynamic length of the train head is between 7.63 and 7.64 m, which differs from the physical length of a realistic train with 12.00 m. Along the noise barrier from bottom to top, the ΔP values on the inner surface decrease with the increase of height, while these values on the outer surface increase. Taking the wind direction into account, the ΔP values are a little higher when the high-speed trains run against the wind direction.
1. Introduction In order to use the land economically, reduce the influence of highspeed railways on the environment and avoid the occupation of fertile land, most of high-speed railway infrastructure scenarios in China are built in the form of bridges. Furthermore, noise barriers are always installed on both sides of the high-speed railway bridge to reduce noise pollution for residents living around the lines (Long et al, 2010). When a high-speed train passes close to the noise barriers along the track, it would exacerbate the turbulence intensity of surrounding air and cause unsteady transient aerodynamic pressure on the barriers' surfaces. In turn this forms a transient pressure impact that reaches positive and negative pressure peaks so quickly within tens of milliseconds (Baker, 2010; Tian, 2007; Schetz, 2001). These train-induced fluctuating pressure amplitudes on both sides of the noise barriers increase rapidly with the square of the train speed (Tian, 2007; PECþS, 2007). And the dilatational and compression effects of fluctuating forces lead to structure fatigue failure at the joints of noise barriers (Zhu and Cheng, 2011; Chen et al., 2011). In 2003, an event related on the noise barriers along the Cologne-Frankfurt
high-speed line took place in Germany due to the influence of these fluctuating loads, and it cost approximately 30 million Euros to reconstruct and repair the barriers (PECþS, 2007). Therefore, to guarantee the operational safety of railway lines, it is significantly important to investigate the fluctuating loads that act on the noise barriers as the high-speed train passes. Some previous studies have been conducted to research the effects of fluctuating pressure on the noise barriers and buildings surrounding the railway line during the passage of high-speed trains. Most of these are carried out in United Kingdom, Germany, Japan and USA. In United Kingdom, Baker (2010, 2014) studied the changes in the regularity of fluctuating forces suffered by the surrounding buildings when three different shapes of trains passed through the noise barriers, bridges, platform roofing and station platforms. This was performed by using a moving model with the scale of 1/25. The research findings are applied to complement the EU railway standard and the 2013 EN (CEN European Standard, 2013). These standards provide recommended loads based on the fluctuating forces on both sides and at the top of the buildings along the railway line (CEN European Standard, 2013). In Germany, the noise
* Corresponding author. Key Laboratory of Traffic Safety on Track of Ministry of Education, School of Traffic & Transportation Engineering, Central South University, Changsha 410075, China. E-mail address: [email protected] (J. Zhang). https://doi.org/10.1016/j.jweia.2018.04.017 Received 29 January 2018; Received in revised form 10 March 2018; Accepted 15 April 2018 0167-6105/© 2018 Elsevier Ltd. All rights reserved.
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
2. Experiment set-up
barriers dynamic response tests were carried out in the Nuremberg–Ingolstadt high-speed railway line. In this test, the fluctuating pressure, dynamic deformation and natural frequency caused by different train types, train speeds and noise barrier types were obtained. At last, they clearly presented a new formula to calculate the fluctuating wind loads based on the measurements of fluctuating pressure on the high-speed railway noise barriers (PECþS, 2007). In Japan, Tokunaga et al. (2016) studied the dynamic response evaluation of noise barriers during the passage of high-speed trains based on field tests and numerical simulations. The effects of some factors on the noise barrier response were quantified. At last, two practical design methods for evaluating the dynamic response of noise barriers were proposed. In USA, the Federal Railroad Administration of U.S. Department of Transportation used field measurements to test the effect of fluctuating forces on the stable double-decker container trains when a high-speed train passed. This force was determined from a vehicle dynamics model, and the vehicle response under the action of the fluctuating force was obtained (Lee, 1999; Holmes and Schroeder, 2002). Carassale and Brunenghi (2013) studied the dynamic response of buildings near the track due to the fluctuating pressure, and developed a non-dimensional mathematical model to predict the dynamic response of the trackside structure subject to the fluctuating pressure caused by passing trains. Additionally, in China some researchers also investigated the aerodynamic pressure loads acting on the noise barriers and surrounding buildings along high-speed train railway lines using tests and simulations (Chen et al., 2011; Zhang et al., 2009). Yang et al. (2015) used the numerical simulations and full scale train tests to study the fluctuating pressure characteristics on an overhead bridge, they observed that the frequency of the fluctuating force was mainly concentrated in the range of 0 Hz–2.5 Hz. Zhou et al. (2014) used a moving model at the ratio of 1/20 to study the fluctuating force acting on the shielding door and the roof of the station platform, and the effect of train speeds, passing position of trains and other parameters of the fluctuating force were obtained. They found that when two trains passing by each other at the speed of 350 km/h in the station, the air compressibility should be considered. Zhang et al. (2009) studied the distribution of fluctuating pressure acting on the sound barrier of a railway bridge with collision walls and box girder flange plates when a high-speed train passed, and the distribution characteristics of fluctuating loads along the height of the noise barrier were obtained. They found that the peak values on the noise barrier and the box girder flange plate are proportional to the square of the train speed. At present, the literature published are mainly focused on the fluctuating pressure on the noise barrier when the train speed is lower than 350 km/h. However, the spatial distribution of fluctuating pressure and the influence factors of fluctuating force are not yet clear when the train speed is higher than 350 km/h. At the same time, the complete and efficient measurement and evaluation methods on the fluctuating pressure of high-speed railway noise barriers haven't been established in China. In this study, a field experiment is conducted to obtain the distribution law of fluctuating pressure on the noise barriers caused by the passage of high-speed trains at the speed varying from 250 km/h to 380 km/h. The results are used to provide data support for the study on the dynamic response evaluation of high-speed railway noise barriers under fluctuating pressure. At the same time, the results are used to contribute to the data reference for revising and improving the relevant standards for the noise barriers along high-speed railways (Chinese National Railway Standard, 2014). However, due to the limitation of the field tests, it is difficult to obtain sufficient data about the effect of environmental wind on the fluctuating pressure on the noise barriers. In the next work, a further analysis of numerical simulations on the influence of environmental wind directions and wind speeds on the fluctuating pressure should be carried out.
2.1. Instrumentation The test section was set up in the middle of a 2.5 km long bridge with noise barriers on both sides along the Beijing-Shanghai high-speed railway. A number of air pressure sensors fixed on the windward side, leeward side, and the top surface of these noise barriers were used to test the fluctuating pressure on the barriers. There were 16 pressure measurement points, of which 12 measurement points were set on the noise barrier close to the tracks (inner surface), and the other points were set on the noise barrier away from the tracks (outer surface). Fig. 1 shows the schematic diagram of the test set-up. The height of the noise barrier was 2.15 m from its top to the rail, and the specific installation section is shown in Fig. 2. The monitoring points in tests are shown in Fig. 3. In this system, LL-250 type pressure sensors produced by the Kulite Company in USA were used to measure the fluctuating forces on noise barriers, and the test range is 15 psi. The thicknesses of these sensors are very thin to minimize the interference on the flow field around barriers. Thus, the results are obtained with high precision. Some approaches were adopted to reduce the measurement noise, such as the use of pressure sensors and testing equipment with stronger anti-interference performance, the shielded signal lines and the UPS power supply. A Stalker Speed Sensor (S3) Police Option radar, seeing Fig. 4, with its measurement accuracy of 1 mph, was placed near the measurement points on the noise barrier to monitor the train speed as a CRH380A EMU passed it. The width of CRH380A EMU is 3.38 m and the height is 3.70 m. The total length of a 16 car-grounding EMU is 403 m and the length of an 8 car-grounding EMU is 203 m. According to the CEN European Standard (2013), in the test a sensor (FLUKE 975 was used in this test) is required to be installed along the railway and is used to monitor the temperature and humidity when the train passes the test position. The temperature range of the sensor is 20–50 C, and the range of relative humidity is 10%–90% RH. Wind speed and wind direction sensors were also installed in the upstream of the bridge to check the wind speeds and directions, as shown in Fig. 5. The XFY3 type wind speed and direction sensors with the range of 0–100 m/s were adopted in the test. The precision of the sensor is less than or equal to 0.5 m/s when the wind speed is 0–10 m/s. However, when the wind speed reaches 10–100 m/s, the precision is changed as less than or equal to 5% of the measured value. As to the wind direction of the sensor, its range is 0–360 and the precision is 3 . These data are used to investigate the impact of environmental wind speeds on the fluctuating pressure distributions on the surfaces of noise barriers. The wind speed sensors were fixed at a height of 3 m above the ground and 30 m far from the bridge, and this height was lower than the bridge height that is 10.5 m. Therefore, the wind speeds measured by the wind speed sensors were less than the wind speed at the noise barrier position. In the test, the sampling frequency of wind speed sensor is 4 Hz and the wind speed is the mean wind speed in the 15s interval before the train nose passes. This data processing method can are referred to that in the CEN European Standard (2013). Due to the limitation of test conditions, the wind data are recorded at the height of 3 m from the ground. In order to obtain the wind speed at the bridge height, a power law is used (Frost, 1948; Wu, 2014). Generally, the wind profile of the atmospheric boundary layer (ABL) (from surface to around 200 m) is logarithmic in nature and is best approximated using the log wind profile equation that accounts for surface roughness and atmospheric stability (Cook, 1985). However, sometimes the surface roughness or stability information is not available, so the wind profile power law relationship is often used as a substitute for the log wind profile. In this section, a power law is used to describe the
158
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 1. Layout of measurement points on the noise barrier.
Fig. 2. High-speed railway bridge with noise barriers on both sides: (a) Cross-sectional view, (b) Top view.
Fig. 3. Measurement points on the barriers in tests: (a) Inner surface, (b) Outer surface.
u ¼ ur ðz=zr Þα
vertical distribution of horizontal mean wind speeds, which is: u=ur ¼ ðz=zr Þα
(1)
(2)
where u is the wind speed at height z, and ur is the known wind speed at a reference height zr ¼ 10 m above the ground based on the Chinese Standard, 2012—Load code for the design of building structures (GB50009-2012). The exponent (α) is an empirically derived coefficient
In order to estimate the wind speed at a certain height z, the relationship would be arranged to:
159
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 6. Schematic diagram of data definitions.
and direction in the tests and to obtain the influence of environmental wind on the fluctuating pressure on the noise barriers.
Fig. 4. Arrangement of the velocity radar.
2.3. Data processing To deal with those test data, a sampling frequency 1 kHz and a sixth order Butterworth low-pass filter 40 Hz are utilized. The data definitions diagram is shown in Fig. 6. ΔP indicates the peak-to-peak pressure value with the largest difference between the maximum positive peak (denoted by Pmax), and the maximum negative peak (denoted by Pmin). The results of several runs at the same speed grade were analysed to determine the reliability and repeatability of the test system. The No. 1 measurement point, as shown in Fig. 1, was selected as an example to analyse this system's repeatability, when a 16 cars marshalling CRH380A EMU repeatedly passed it at the speed of 380 km/h. There were three operational runs: as the train ran from north to south, the wind direction was from west to east and the wind speeds were 0.78 m/s, 0.85 m/s and 1.06 m/s. The time history curves of pressure in the 3 runs are illustrated in Fig. 7, and the test results are presented in Table 2. During data processing, if the train speed difference is in the range of approximately 2 km/h to a predetermined speed level, the data is considered to be valid at this level. With respect to the influence of environmental wind on the pressure, when the wind speed is below 2 m/ s, the influence is ignored. At each train speed level, experiments are repeated at least three runs, and then the fluctuating pressure data at each measurement point are obtained by using the ensemble average method. Fig. 7 indicates that the pressure curves of the No. 1 measurement point in 3 runs are basically identical. The pressure changing curves due to the passage of the train head and tail are substantially coincident, and the curves are slightly different when the middle cars pass, due to a small difference in the train speed and environmental wind. Since the environmental wind direction and wind speed are not
Fig. 5. Arrangement of the wind speed sensor.
that varies dependent on the stability of the atmosphere. Because the test point is located at the flat terrain and open land cover conditions, for this terrain, α is approximately 0.15 (GB50009-2012). So in this paper, α ¼ 0.15 is used.
2.2. Test cases As it is known, field tests cost large amounts of resources, leading to limited runs for each train speed, so it is very difficult to obtain lots of data, especially for those in the environmental wind conditions. In this full-scale test, the test runs and corresponding train speeds are listed in Table 1. The train passes the test section on the near rail line 29 runs and on the far rail line 8 runs. Generally, the number of total runs is enough for the field tests, because the environmental wind is always not taken into account, except for the tests in strong wind regions. According to the climate data along the railway in the past years, there are some days with slight wind. Therefore, sensors were set up to monitor the wind speed Table 1 Test number at different train speeds. Train speed (km/h)
Runs Near rail
Far rail
250 300 350 380
4 5 8 12
0 4 4 0
Fig. 7. Time history curves of pressure in the 3 runs. 160
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Table 2 Repeatability of measurement results. Case
Pmax (Pa)
Pmin (Pa)
ΔP (Pa)
Standard deviation (Pa)
Relative standard deviation (%)
The first run The second run The third run
883 903 881
680 650 657
1 563 1 554 1 538
10.34
0.67
However, it still fluctuates in a small limited range around the ambient pressure due to the complex structures of bogies and inter-carriage gaps (CEN European Standard, 2013; Hemida et al., 2014; Huang et al., 2017; Niu et al., 2017; Zhang et al., 2018). When the train tail passes by the measurement point, the pressure first decreases and then increases instantly, and a pressure wave turns up with a negative peak followed by a positive peak. Overall, these typical pressure variations come from the train head and tail, and the pressure caused by the passage of the middle cars is much smaller. The most severe variation of pressure is caused by the passage of the train head, as a result the peak-to-peak values of the head wave are much larger than these of the tail wave, as reported in Zhou et al. (2014), CEN European Standard (2013), Hemida et al. (2014), Huang et al. (2017), Niu et al. (2017), Zhang et al. (2018). Note that the peak-to-peak pressure values on the inner surface of the barrier are much larger than those on the outer surface at the same height, as the air is posed on the points in different impact ways. For the point on the inner surface, the air is pushed by the train directly acts on it, while for the outer one the pressure variation is induced by the air from the railways in an indirect approach. The maximum positive peak and negative peak of the fluctuating pressure on the inner surface of the barrier occurred when the train head passed the measurement point, and the absolute value of the maximum positive peak of the pressure is larger than the maximum negative peak. In contrast, for the tail wave, the absolute value of the negative peak is larger than the positive peak. Therefore, the starting and end points, i.e., the noses, play significant roles in these pressure waves, which contributes to a larger absolute value of pressure. The phenomenon is also found in Baker et al. (2014). In the wake of the high-speed train, the flow is more turbulent and complex, and two counter-rotating vortices are shown (Bell et al., 2016, 2017; Zhang et al., 2016). As a result, more fluctuations from the flow structures are expected to cause the variations of the pressure at the monitoring points after the train travel away, as remarked in red circles in Fig. 8. Additionally, as the previous discussion, this fluctuation of the inner point is much stronger than that of the outer surface, which is also reported by Baker (2014a,b).
repetitive along the railway and the environmental wind speed is less than 2 m/s, the test data in Table 2 present high accuracy with the relative standard deviation of 0.67%. 3. Test results and analysis 3.1. Regularity analysis of the fluctuating pressure on the barrier 3.1.1. Characteristic curve analysis Fig. 8 shows the time history curves of fluctuating pressure on inner and outer surfaces of the noise barrier when a 16 car marshalling CRH380A EMU passes at the speed of 380 km/h. This typical curve has been reported in CEN European Standard, 2013 and Zhou et al. (2014). The fluctuating pressure is generated when the train passes close to the noise barrier, which is a dynamic process changing over time. It is interesting to note that this history curve is similar to that the pressure sweeps along a line that is parallel to the length of the train with the same distance (CEN European Standard, 2013; Hemida et al., 2014; Huang et al., 2017; Niu et al., 2017; Zhang et al., 2018). When the train head approaches the measurement point, the pressure starts to climb up to the highest. This is because the train pushes air towards the inner surface of the barrier. The positive peak occurs at the arrival of the train nose. After that, the pressure drops rapidly to a negative peak. Therefore, a complete pressure wave is generated by the train head. After the train head passes the measurement point, as it is found in (CEN European Standard, 2013; Zhou et al., 2014), the pressure fluctuation at the observation point considerably slows down, as the cross section of the train is uniform.
3.1.2. Relationship curves between the fluctuating pressure and train speed Fig. 9 shows the fluctuating pressure curves of the No.1 measurement point (Inner surface) and the No. 8 measurement point (Outer surface) on the barrier, when a 16 car marshalling CRH380A EMU passes close to the noise barrier at different speeds. Here to show the differences of fluctuating pressure curves at different train speeds, it is supposed that the tail nose of the train leaves the points at the same time, because in this curve the most severe variation of pressure comes from the train head passing. Clearly a higher train speed contributes to a larger peak-to-peak pressure of the head wave no matter where the monitoring points are. Meanwhile, the fluctuations induced by the middle cars also enhance with the train speed, seeing Fig. 9. Fig. 10 shows the law curves of ΔP, defined as the difference of Pmax to Pmin in Fig. 6, on the inner and outer surfaces of the barrier with changes of the train speed. According to the research conducted by Tokunaga et al. (2016), Tian (2007) and Zhou et al. (2014), the ΔP values of the head wave are approximately proportional to the square of the train speed. Here, this law is clearly shown on the inner surface of the barrier. The pressure amplitudes on the outer surface at different locations also show power laws with respect to the train speed. However, these exponents that present some differences, as shown in Fig. 10 (b), reduce with decrease in the height of monitoring points. Compared to the pressure
Fig. 8. Time history curves of fluctuating pressure on measurement points: (a) Inner surface, (b) Outer surface. 161
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 9. Fluctuating pressure curves of different measurement points: (a) Point 1 (b) Point 8.
Fig. 10. Relationship curves between ΔP and train speeds: (a) Inner surface, (b) Outer surface.
amplitudes on the inner surface of the barrier, seeing Fig. 10(a), these amplitudes are so small that may be easily affected by the measurement accuracy of the sensors and environmental wind. Therefore, roughly, the pressure amplitudes are proportional to the square of the train speed no matter on the inner and outer surfaces of the barrier. To analyse the train speed impact on the fluctuating pressure on barrier surfaces, a peak-to-peak time interval, denoted by Δt, is the time from Pmax to Pmin of the head wave is defined according to CEN European Standard (2013). Where, Ln is the length of the train head and vtr is the train speed. Table 3 shows the Δt, when a 16 grouping-car CRH380A EMU passes the No.1 measurement point at different train speeds. Δt Ln = vtr
Table 3 Peak-to-peak time intervals. Train running speed(km/h)
250
300
350
380
Δt (s) Ln (m) from tests Ln (m) from a realistic train
0.1099 7.6319 12.0000
0.0917 7.6389
0.0786 7.6384
0.0723 7.6317
distance from the far railway line is 8.40 m ΔPnear is defined as the ΔP at the points when the train runs on the near railway line, while ΔPfar is the ΔP at the points when the train runs on the far railway line. The general observation is that when the lateral distance between the train and the barrier is smaller, the train induced pressure on the point fluctuates largely, especially for the head wave, as shown in Fig. 12(a). This is because the energy of the wave is diffused in the propagation. It is similar to these shown in Fig. 12(b) that the ΔP reduces with the increase in height of monitoring points. However, due to the same scale for the ΔPnear and ΔPfar, it is a little difficult to see the decrease trend for ΔPfar. Overall, the ratios ofΔPnear to ΔPfar for these 6 monitoring points are in a range of 0.245–0.262.
(3)
As the train speed increases, the peak-to-peak time interval Δt decreases gradually and the corresponding slope becomes increasingly steep. According to formula (3), the train head length is obtained from Ln Δt * vtr, as shown in Fig. 11(a). Note that this length actually corresponds to the length from the front tip of the nose to the transition curve position of the cabin window instead of the 12 m long train head of a realistic high-speed train, seeing Fig. 11(b). Furthermore, the position of the maximum negative pressure on the train body surface in the test is essentially consistent with that previously reported by Yang et al. (2015) when a high-speed train passes on the bridge. The result shows that there is a debate about the definition of the length Ln of the train head in aerodynamics, and the parameters require to be redefined.
3.1.4. Comparison with other sources of data From Section 3.1.1, it is shown that the absolute value of the positive peak of the fluctuating pressure is larger than that of the negative peak. So in this section, the positive head pulse pressure coefficients of the test results are compared with other experimental data and some data from empirical equations, as shown in Fig. 13. The measurements were conducted by Tokunaga et al. (2016) and Horie and Sugiyama (1986). For the equations, one is included in the 2013 EN (CEN European Standard, 2013) that is Cp ¼ 2.5/(Yþ0.25)2þ0.02, where the Y is the distance from the track centre, which is greater than 2.3 m. The other equation is from a literature published by Baker et al. (2014), as follows: Cp ¼ 6/(Yþ1.75)2. When the distance between the noise barrier and the track centre is
3.1.3. Comparative analysis of the fluctuating pressure and the train running lines Fig. 12 (a) shows the time history curves of fluctuating pressure at the No. 1 measurement point on the inner surface of the noise barrier when a train runs on the near and far rails at the speed of 350 km/h, respectively. A comparison of ΔP at points No. 1–6 is found in Fig. 12(b). The distance of the inner surface of the barrier from the centreline of the track is 3.40 m when the CRH380A EMU runs on the near railway line, and the 162
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 11. Length of the train head: (a) Test length, (b) Actual length.
pressures and forces on noise barrier are underestimated. Therefore, it seems better to adopt the equation from Baker et al. (2014) to estimate the transient aerodynamic pressures on the noise barrier. Additionally, the experimental results of the transient aerodynamic pressure coefficients in Japan are less than the calculated data using the equation in the 2013 EN (CEN European Standard, 2013). Based on above analysis, although the empirical equations can be used to estimate the transient aerodynamic pressures on the noise barriers in China, these data still show slight differences compared with the data in the present work, which is not in accordance with the actual situation of China high-speed railways. Therefore, it is very important to obtain those data directly coming from China high-speed railways. 3.2. Distribution law of ΔP along the vertical direction on the noise barrier Fig. 14 shows the changing trends of ΔP on the inner and outer surfaces of the noise barrier along the height direction when a 16 car marshalling CRH380A EMU passes the test points on the noise barrier at different speeds. As illustrated in Fig. 14, the ΔP values on the inner surface of the noise barrier decrease with the increase in the height of measurement points from bottom to top. At the highest point they are approximately 74.5%– 81.9% of these at the bottom when the CRH380A EMU passes the test section at different speeds. The ΔP distributions on the noise barrier shows a slight change from the centre to the bottom, and the values at the centre are approximately 92.4%–95.2% of these at the bottom. The phenomenon is also found in the report (PECþS, 2007) and the height of the noise barrier is 4.0 m. This is mainly because the effective airflow area at the bottom of the noise barrier is less than that at the top, and the airflow velocity at the bottom is higher than that at the top. Furthermore, the direction of airflow shows an outflow trend to the outer surface of the noise barrier, and thus the pressure at the bottom of the noise barriers is greater than that at the top. The ΔP values at the bottom of the outer surface of the noise barrier are the smallest along the height direction. Then they gradually increase with the height. However, at the bottom of the inner surface of the noise barriers they are the largest. Due to the shielding effect of noise barriers, the fluctuating pressure generates when the train passes the measurement points and reaches at the top of the outer surface of the noise barriers, then spread to the bottom of the outer surface down along the noise barriers. This result in the ΔP values of the measurement points at the top the barriers exceeding that at the bottom. As the ΔP values on the inner surface close to the bottom of noise barriers are the largest, it is required to design a more appropriate bottom of the noise barrier.
Fig. 12. Pressure change curves and peak-to-peak comparison: (a) Pressure variation at point 1, (b) Peak-to-peak comparison of fluctuating pressure at points 1-6.
3.3. Fluctuating pressure distribution law on the noise barriers along the train running direction
Fig. 13. Comparison of pressure coefficients with previous work.
3.40 m, the full scale pressure coefficients on the inner surface at different train speeds are slightly larger than those calculated data using the recommended equation in the 2013 EN (CEN European Standard, 2013), whereas these coefficients are smaller than those calculated data using the proposed equation from Baker et al. (2014). If the equation in the 2013 EN (CEN European Standard, 2013) is used to calculate the pressure coefficient of the head pulse, the transient aerodynamic
Fig. 15 shows the changing curves and ΔP comparison chart at different measurement points which are fixed on the inner surface of the noise barrier at the same height but different longitudinal positions, when the 16 car marshalling CRH380A EMU passes the test points at different speeds. Although those points are at the same height, there are some distances in the train running direction among those points, especially for the point 16 with a distance of 150 m from point 4 which is very 163
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 16. Time history curves of fluctuating pressure and peak-to-peak comparison at the measurement points on the inner surface: (a) Time history curves of fluctuating pressure, (b) Peak-to-peak comparison chart of fluctuating pressure.
Fig. 14. Changing trends of peak-to-peak pressure values along the height direction: (a) Inner surface, (b) Outer surface.
close to points 12 and 15, as shown in Fig. 1. As a result, the time when the train head arrives those points is different, as shown in Fig. 15(a). Since the heights are the same for all points, according to CEN European Standard (2013), the time history curves of all points should be very similar. In fact, from the measurement, the ΔP values of monitoring points along the train running direction are almost the same when the points are located in the middle of the noise barrier. Additionally, when a 16 car marshalling CRH380A EMU passes the test points, the ratios of the maximum to the minimum of the peak-to-peak pressure values at different points with the same height are between 101.0% and 101.7%. This error comes from the measurement error of the test system and the relative height between the bottom of the barrier and the rail top. 3.4. Influence of different train marshalling lengths on the fluctuating pressure Fig. 16 shows the time history curves of fluctuating pressure, Fig. 16(a), and the ΔP values comparison chart, Fig. 16(b), when a CRH380A EMU with 2 types of marshalling (16 car-grouping and 8 cargrouping) passes close to the noise barrier at 380 km/h and the environmental wind speed is less than 2 m/s. The previous analysis indicates that the fluctuating pressure distribution on the surface of the noise barriers mainly depends on the train speed, the streamlined head and the couplings (CEN European Standard, 2013; Baker, 2014a,b). In the present work, the streamlined head and the train speed are the same and there is no coupling in the 16 car-grouping CRH380A EMU, so the ΔP values at the identical point should be basically the same. In fact, when the CRH380A EMU with the 16 car and 8 car marshalling types passes the measurement points on the noise barrier at the same speed, the head wave and the ΔP values at the same measurement point are basically the same, and the largest difference between
Fig. 15. Time history curves of fluctuating pressure and peak-to-peak comparison chart at different measurement points with the same height: (a) Time history curves of fluctuating pressure, (b) Peak-to-peak comparison chart of fluctuating pressure.
164
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Table 4 Influence of environmental wind on the peak-to-peak pressure values. Running direction
Environmental wind speed (m/s)
Environmental wind direction
ΔP (Pa)
From From From From From
8.06 5.86 1.44 5.18 6.53
Due north Due north Due north Due north Due north
1 383 1 434 1 599 1 679 1764
north to south north to south south to north south to north south to north
them is only 1.74%, which further confirms the fluctuating pressure distribution on the surface of the noise barriers and the ΔP mainly depends on the train speed, the streamlined head and the couplings.
pressure values is 27.6% larger than the minimum, when the environmental wind speed changes within the range of 0–8.06 m/s, and the 16 car marshalling CRH380A EMU runs on the line close to the noise barrier in different directions at the speed of 380 km/h. Given that the environmental wind velocity is generally small in the tests and the direction is relatively fixed, this study only analyses the impact of the environmental wind velocity on the fluctuating pressure. Hence, an in-depth analysis of the environmental wind direction on the influence of fluctuating pressure is not performed, which are continued to study by numerical simulations in the next work.
3.5. Influence of environmental wind on the fluctuating pressure on noise barriers In the tests, 29 runs on the near rail line and 8 runs on the far rail line were carried out, respectively, as listed in Table 1. The number of total runs is sufficient for a conventional field test, and 5 runs come across environmental wind if the wind speeds recorded by the wind speed and direction sensors as shown in Fig. 5 are not less than 1 m/s. Although the samples seem not enough to be used for concluding the influence laws of environmental wind on the fluctuating pressure on noise barriers, general relationships between the wind velocity, wind direction and the peak-to-peak pressure measured are obtained. Also those data are used to validate against the numerical simulations in the next work. The ΔP values at the measurement point 1 as a 16 car marshalling CRH380A EMU passes close to the noise barrier at the speed of 380 km/h under different environmental wind speeds and directions as shown in Table 4. Fig. 17 shows the time history curves of fluctuating pressure at point 1 on the inner surface of the noise barrier under different wind environments, when the 16 car marshalling CRH380A EMU passes the point at the speed of 380 km/h. The peak-to-peak pressure values on the noise barrier surfaces are significantly affected by the environmental wind. When the CRH380A EMU runs against the wind direction, the peakto-peak pressure values at the measurement point on the inner surface of the noise barrier are larger than these when the train runs along the wind direction. This is because the relative airflow speed at the measurement points on the noise barrier surfaces are superimposed by both the flow velocity induced by the CRH380A EMU and wind velocity component along the line direction in the existence of environmental wind. In other words, when the train runs against the wind, the relative wind speed at the measurement point is the sum of the flow velocity induced by the CRH380A EMU and wind velocity component along the line direction. In contrast, the relative wind speed is the difference between the flow velocity induced by the CRH380A EMU and wind velocity component along the line direction. Moreover, the maximum in these peak-to-peak
4. Conclusions A full scale field test has been conducted to obtain the variations of pressure on a 2.15 m high bridge noise barrier induced by the passage of a CRH380A EMU high-speed train at the speed of 250 km/h ~380 km/h using LL-250 type pressure sensors which are produced by the Kulite Company in USA. A Stalker Speed Sensor (S3) Police Option radar with high accuracy is used to measure the train speed when the train passes by the testing region, and XFY3 type wind speed and direction sensors are installed in the upstream of the bridge to check the wind speeds and directions. The time history curves of fluctuating pressure on the inner and outer surfaces of the barrier, especially the peak-to-peak pressures caused by the passage of the train head and tail, as found by previous research in scaled, numerical and full scale studies. The positive head pulse pressure coefficients of the test results show slight differences, compared with other experimental data and some data from empirical equations. Further, train noses have significant effects on the first part of the head and tail pressure waves, i.e., a larger absolute value for the first peak in both waves. The ΔP distributions on the inner and outer surfaces of the barrier are not identical; indicating the pressure on the outer surfaces is induced by the air from the railways in an indirect approach and is easily affected by the ambient environment. Due to two counterrotating vortices in the wake of the train tail, the pressure on the barrier continues fluctuating after the train leaves. The ΔP values of the head wave are approximately proportional to the square of the train speed, as reported in previous investigations. With the increase of the train speed, the peak-to-peak time intervals of ΔP decrease gradually and the slopes become increasingly steep. The findings from the tests indicate that for a CRH380A EMU the aerodynamic length of the train head is between 7.63 and 7.64 m. This is different from the physical length of the realistic train head with 12.00 m. Thus, there is a discrepancy with respect to the definition of the train head length (denoted by Ln), and the parameters need to be redefined. The distance of the barrier from the train symmetry plane determines the ΔP value. When the lateral distance between the train and the barrier is smaller, the train induced pressure on the point fluctuates largely, especially for the head wave. Generally, the ratios of ΔPnear to ΔPfar are in a range of 0.245–0.262. The train marshalling length does show much effect on the fluctuating pressure time history curve and ΔP, if there are no couplings in a long marshalling train. Along the noise barrier height, the ΔP values on the inner surface of the noise barrier decrease with the increase in the height of measurement points. Additionally, the ΔP values on the outer surface of the noise barrier increase with the height increasing. Taking the wind speed and direction into account, when the environmental wind speed changes
Fig. 17. Curves of fluctuating pressure at point 1 on the inner surface of the noise barrier under different wind environments. 165
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
within the range of 0–8.06 m/s and the 16 car marshalling CRH380A EMU runs on the line close to the noise barrier in different directions at the speed of 380 km/h, the maximum in these ΔP values is 27.6% larger than the minimum. Overall, the results in this paper can provide data support for the study on the dynamic response evaluation of high-speed railway noise barriers under the fluctuating loads and also contribute to the data reference for revising and improving the relevant standards for the noise barriers along high-speed railways.
Chen, X.D., Li, S.D., Wang, Z.H., 2011. Numerical simulation study on high-speed train induced impulsive pressure on railway noise barrier based on ALE. J. China Railw. Soc. 33 (12), 21–26 (in Chinese). Chinese Standard, 2012. Load Code for the Design of Building Structures. GB50009—2012. Cook, N.J., 1985. The Designer's Guide to Wind Loading of Building Structures: Part 1. Butterworths. Frost, R., 1948. Atmospheric turbulence. Q. J. R. Meteorol. Soc. 74 (321–322), 316–338. Hemida, H., Baker, C.J., Gao, G., 2014. The calculation of train slipstreams using largeeddy simulation. Proc. Inst. Mech. Eng. - Part F J. Rail Rapid Transit 228 (1), 25–36. Holmes, S., Schroeder, M., 2002. Aerodynamic Effects of High-speed Passenger Trains on Other Trains. DOT/FRA/ORD-01/12. Horie, A., Sugiyama, T., 1986. Field Test of Train Draft on Tohoku Shinkansen. Technical Report A-86,Railway Technical Research pre-Report. Huang, S., Hemida, H., Yang, M.Z., 2017. Numerical calculation of the slipstream generated by a CRH2 high-speed train. Proc. Inst. Mech. Eng., Part F: J. Rail. Rap. Tran 230, 103–116. Lee, H., 1999. Assessment of Potential Aerodynamic Effect on Personnel and Equipment in Proximity to High-speed Trains Operations. DOT/FRA/ORD-99/11. Long, L.P., Zhao, L.B., Liu, L.D., 2010. Research on the air turbulent force loading on noise barrier caused by train. Eng. Mech. 27 (3), 246–250 (in Chinese). Chinese National Railway Standard, 2014. Code for Design of High Speed Railway, TB 10621. Niu, J.Q., Zhou, D., Liang, X.F., 2017. Numerical simulation of the effects of obstacle deflectors on the aerodynamic performance of stationary high-speed trains at two yaw angles. Proc. Inst. Mech. Eng. - Part F J. Rail Rapid Transit 232. Schetz, J.A., 2001. Aerodynamics of high-speed trains. Annu. Rev. Fluid Mech. 53 (2), 371–414. Tian, H.Q., 2007. Train Aerodynamics. China Railway Publishing House (In Chinese). Tokunaga, M., Sogabe, M., Santo, T., Ono, K., 2016. Dynamic response evaluation of tall noise barrier on high speed railway structures. J. Sound Vib. 366, 293–308. Wu, Y., 2014. Wind Engineering and Structure Wind-resistant Design. Harbin Institute of Technology Publishing House (In Chinese). Yang, N., Zheng, X.K., Zhang, J., Law, S.S., Yang, Q.S., 2015. Experimental and numerical studies on aerodynamic loads on an overhead bridge due to passage of high-speed train. J. Wind Eng. Ind. Aerod. 140 (10), 19–33. Zhang, J.W., Lü, J.P., Tu, Y.M., Kong, C.W., Ge, W.J., 2009. Numerical analysis of highspeed train induced impulsive pressure on railway bridge's noise barrier and its flange Plate with collision-preventing wall. China Railw. Sci. 30 (3), 27–32 (In Chinese). Zhang, J., Li, J.J., Tian, H.Q., Gao, G.J., Sheridan, J., 2016. Impact of ground and wheel boundary conditions on numerical simulation of the high-speed train aerodynamic performance. J. Fluid Struct. 61, 249–261. Zhang, J., Wang, J.B., Wang, Q.X., Xiong, X.H., Gao, G.J., 2018. A study of the influence of bogie cut outs' angles on the aerodynamic performance of a high-speed train. J. Wind Eng. Ind. Aerod. 175, 153–168. Zhou, D., Tian, H.Q., Zhang, J., Yang, M.Z., 2014. Pressure transients induced by a high speed train passing through a station. J. Wind Eng. Ind. Aerod. 135, 1–9. Zhu, Z.Q., Cheng, Z.Q., 2011. Numerical simulation and experimental verification of dynamic pressure on noise barriers of high speed railways. Railway standard design. 11, 77–80 (in Chinese).
Acknowledgments This work was supported by thanks the National key R & D program of China (Grant No. 2016YFB1200506-03), the National Science Foundation of China (Grant Nos. 51205418, 51605044 and U1334205), the Science Foundation of Hunan Province of China (Grant No. 2016jj3004) and the Science and Technology Research Programme of China Railway Corporation (Grant No.2016T004-C). References Baker, C.J., Jordan, S., Gilbert, T., Quinn, A., Sterling, M., Johnson, T., Lane, J., 2014. Transient aerodynamic pressures and forces on trackside and overhead structures due to passing trains. Part1 Model scale experiments. Part2 Standards applications. Proc IMechE Part F: J.Rail Rapid Transit 228 (1), 37–70. Planning Engineering Consulting þ Services China Ltd, 2007. Consultation Report of the Noise Barriers in Chinese Railway Passenger Dedicated Line. Baker, C.J., 2010. The flow around high speed trains. J. Wind Eng. Ind. Aerod. 98, 277–298. Baker, C.J., 2014a. A review of train aerodynamics part1-fundamentals. Aeronaut. J. 118 (1201), 201–228. Baker, C.J., 2014b. A review of train aerodynamics part1-applications. Aeronaut. J. 118 (1201), 345–382. Bell, J.R., Burton, D., Thompson, M.C., Herbst, A.H., Sheridan, J., 2016. Flow topology and unsteady features of the wake of a generic high-speed train. J. Fluid Struct. 61, 168–183. Bell, J.R., Burton, D., Thompson, M.C., Herbst, A.H., Sheridan, J., 2017. A wind-tunnel methodology for assessing the slipstream of high-speed trains. J. Wind Eng. Ind. Aerod. 166, 1–19. Carassale, L., Brunenghi, M.M., 2013. Dynamic response of track side structures due to the aerodynamic effects produced by passing trains. J. Wind Eng. Ind. Aerod. 123, 317–324. CEN European Standard, 2013. Railway applications – aerodynamics. In: Part4: Requirements and Test Procedure for Aerodynamics on Open Track. CEN EN 14067–4.
166
Contents lists available at ScienceDirect
Journal of Wind Engineering & Industrial Aerodynamics journal homepage: www.elsevier.com/locate/jweia
Field study on high-speed train induced fluctuating pressure on a bridge noise barrier Xiao-hui Xiong a, b, c, Ai-hua Li a, b, c, Xi-feng Liang a, b, c, Jie Zhang a, b, c, * a b c
Key Laboratory of Traffic Safety on Track of Ministry of Education, School of Traffic & Transportation Engineering, Central South University, Changsha 410075, China Joint International Research Laboratory of Key Technology for Rail Traffic Safety, Changsha, Hunan 410075, China National & Local Joint Engineering Research Center of Safety Technology for Rail Vehicle, Changsha, Hunan 410075, China
A R T I C L E I N F O
A B S T R A C T
Keywords: Fluctuating pressure Field measurement Noise barrier High-speed train Aerodynamics
The pressure variations induced by a CRH380A EMU on a 2.15 m high bridge noise barrier are investigated in a field measurement. The familiar fluctuating pressure time history curves and the peak-to-peak pressure (ΔP) distributions attributing to the train head on the barrier surfaces are presented. A comparison of the positive head pulse pressure is made between the measurement results and the data calculated by empirical equations in EN 14067-4. Furthermore, the influences of train speed, train running lines, locations of measurement points, train marshalling length and environmental wind speed on ΔP are analysed. The results indicate as the train speed increases, the corresponding time intervals of ΔP decrease gradually, whereas their slopes become increasingly steep. For a CRH380A EMU, the aerodynamic length of the train head is between 7.63 and 7.64 m, which differs from the physical length of a realistic train with 12.00 m. Along the noise barrier from bottom to top, the ΔP values on the inner surface decrease with the increase of height, while these values on the outer surface increase. Taking the wind direction into account, the ΔP values are a little higher when the high-speed trains run against the wind direction.
1. Introduction In order to use the land economically, reduce the influence of highspeed railways on the environment and avoid the occupation of fertile land, most of high-speed railway infrastructure scenarios in China are built in the form of bridges. Furthermore, noise barriers are always installed on both sides of the high-speed railway bridge to reduce noise pollution for residents living around the lines (Long et al, 2010). When a high-speed train passes close to the noise barriers along the track, it would exacerbate the turbulence intensity of surrounding air and cause unsteady transient aerodynamic pressure on the barriers' surfaces. In turn this forms a transient pressure impact that reaches positive and negative pressure peaks so quickly within tens of milliseconds (Baker, 2010; Tian, 2007; Schetz, 2001). These train-induced fluctuating pressure amplitudes on both sides of the noise barriers increase rapidly with the square of the train speed (Tian, 2007; PECþS, 2007). And the dilatational and compression effects of fluctuating forces lead to structure fatigue failure at the joints of noise barriers (Zhu and Cheng, 2011; Chen et al., 2011). In 2003, an event related on the noise barriers along the Cologne-Frankfurt
high-speed line took place in Germany due to the influence of these fluctuating loads, and it cost approximately 30 million Euros to reconstruct and repair the barriers (PECþS, 2007). Therefore, to guarantee the operational safety of railway lines, it is significantly important to investigate the fluctuating loads that act on the noise barriers as the high-speed train passes. Some previous studies have been conducted to research the effects of fluctuating pressure on the noise barriers and buildings surrounding the railway line during the passage of high-speed trains. Most of these are carried out in United Kingdom, Germany, Japan and USA. In United Kingdom, Baker (2010, 2014) studied the changes in the regularity of fluctuating forces suffered by the surrounding buildings when three different shapes of trains passed through the noise barriers, bridges, platform roofing and station platforms. This was performed by using a moving model with the scale of 1/25. The research findings are applied to complement the EU railway standard and the 2013 EN (CEN European Standard, 2013). These standards provide recommended loads based on the fluctuating forces on both sides and at the top of the buildings along the railway line (CEN European Standard, 2013). In Germany, the noise
* Corresponding author. Key Laboratory of Traffic Safety on Track of Ministry of Education, School of Traffic & Transportation Engineering, Central South University, Changsha 410075, China. E-mail address: [email protected] (J. Zhang). https://doi.org/10.1016/j.jweia.2018.04.017 Received 29 January 2018; Received in revised form 10 March 2018; Accepted 15 April 2018 0167-6105/© 2018 Elsevier Ltd. All rights reserved.
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
2. Experiment set-up
barriers dynamic response tests were carried out in the Nuremberg–Ingolstadt high-speed railway line. In this test, the fluctuating pressure, dynamic deformation and natural frequency caused by different train types, train speeds and noise barrier types were obtained. At last, they clearly presented a new formula to calculate the fluctuating wind loads based on the measurements of fluctuating pressure on the high-speed railway noise barriers (PECþS, 2007). In Japan, Tokunaga et al. (2016) studied the dynamic response evaluation of noise barriers during the passage of high-speed trains based on field tests and numerical simulations. The effects of some factors on the noise barrier response were quantified. At last, two practical design methods for evaluating the dynamic response of noise barriers were proposed. In USA, the Federal Railroad Administration of U.S. Department of Transportation used field measurements to test the effect of fluctuating forces on the stable double-decker container trains when a high-speed train passed. This force was determined from a vehicle dynamics model, and the vehicle response under the action of the fluctuating force was obtained (Lee, 1999; Holmes and Schroeder, 2002). Carassale and Brunenghi (2013) studied the dynamic response of buildings near the track due to the fluctuating pressure, and developed a non-dimensional mathematical model to predict the dynamic response of the trackside structure subject to the fluctuating pressure caused by passing trains. Additionally, in China some researchers also investigated the aerodynamic pressure loads acting on the noise barriers and surrounding buildings along high-speed train railway lines using tests and simulations (Chen et al., 2011; Zhang et al., 2009). Yang et al. (2015) used the numerical simulations and full scale train tests to study the fluctuating pressure characteristics on an overhead bridge, they observed that the frequency of the fluctuating force was mainly concentrated in the range of 0 Hz–2.5 Hz. Zhou et al. (2014) used a moving model at the ratio of 1/20 to study the fluctuating force acting on the shielding door and the roof of the station platform, and the effect of train speeds, passing position of trains and other parameters of the fluctuating force were obtained. They found that when two trains passing by each other at the speed of 350 km/h in the station, the air compressibility should be considered. Zhang et al. (2009) studied the distribution of fluctuating pressure acting on the sound barrier of a railway bridge with collision walls and box girder flange plates when a high-speed train passed, and the distribution characteristics of fluctuating loads along the height of the noise barrier were obtained. They found that the peak values on the noise barrier and the box girder flange plate are proportional to the square of the train speed. At present, the literature published are mainly focused on the fluctuating pressure on the noise barrier when the train speed is lower than 350 km/h. However, the spatial distribution of fluctuating pressure and the influence factors of fluctuating force are not yet clear when the train speed is higher than 350 km/h. At the same time, the complete and efficient measurement and evaluation methods on the fluctuating pressure of high-speed railway noise barriers haven't been established in China. In this study, a field experiment is conducted to obtain the distribution law of fluctuating pressure on the noise barriers caused by the passage of high-speed trains at the speed varying from 250 km/h to 380 km/h. The results are used to provide data support for the study on the dynamic response evaluation of high-speed railway noise barriers under fluctuating pressure. At the same time, the results are used to contribute to the data reference for revising and improving the relevant standards for the noise barriers along high-speed railways (Chinese National Railway Standard, 2014). However, due to the limitation of the field tests, it is difficult to obtain sufficient data about the effect of environmental wind on the fluctuating pressure on the noise barriers. In the next work, a further analysis of numerical simulations on the influence of environmental wind directions and wind speeds on the fluctuating pressure should be carried out.
2.1. Instrumentation The test section was set up in the middle of a 2.5 km long bridge with noise barriers on both sides along the Beijing-Shanghai high-speed railway. A number of air pressure sensors fixed on the windward side, leeward side, and the top surface of these noise barriers were used to test the fluctuating pressure on the barriers. There were 16 pressure measurement points, of which 12 measurement points were set on the noise barrier close to the tracks (inner surface), and the other points were set on the noise barrier away from the tracks (outer surface). Fig. 1 shows the schematic diagram of the test set-up. The height of the noise barrier was 2.15 m from its top to the rail, and the specific installation section is shown in Fig. 2. The monitoring points in tests are shown in Fig. 3. In this system, LL-250 type pressure sensors produced by the Kulite Company in USA were used to measure the fluctuating forces on noise barriers, and the test range is 15 psi. The thicknesses of these sensors are very thin to minimize the interference on the flow field around barriers. Thus, the results are obtained with high precision. Some approaches were adopted to reduce the measurement noise, such as the use of pressure sensors and testing equipment with stronger anti-interference performance, the shielded signal lines and the UPS power supply. A Stalker Speed Sensor (S3) Police Option radar, seeing Fig. 4, with its measurement accuracy of 1 mph, was placed near the measurement points on the noise barrier to monitor the train speed as a CRH380A EMU passed it. The width of CRH380A EMU is 3.38 m and the height is 3.70 m. The total length of a 16 car-grounding EMU is 403 m and the length of an 8 car-grounding EMU is 203 m. According to the CEN European Standard (2013), in the test a sensor (FLUKE 975 was used in this test) is required to be installed along the railway and is used to monitor the temperature and humidity when the train passes the test position. The temperature range of the sensor is 20–50 C, and the range of relative humidity is 10%–90% RH. Wind speed and wind direction sensors were also installed in the upstream of the bridge to check the wind speeds and directions, as shown in Fig. 5. The XFY3 type wind speed and direction sensors with the range of 0–100 m/s were adopted in the test. The precision of the sensor is less than or equal to 0.5 m/s when the wind speed is 0–10 m/s. However, when the wind speed reaches 10–100 m/s, the precision is changed as less than or equal to 5% of the measured value. As to the wind direction of the sensor, its range is 0–360 and the precision is 3 . These data are used to investigate the impact of environmental wind speeds on the fluctuating pressure distributions on the surfaces of noise barriers. The wind speed sensors were fixed at a height of 3 m above the ground and 30 m far from the bridge, and this height was lower than the bridge height that is 10.5 m. Therefore, the wind speeds measured by the wind speed sensors were less than the wind speed at the noise barrier position. In the test, the sampling frequency of wind speed sensor is 4 Hz and the wind speed is the mean wind speed in the 15s interval before the train nose passes. This data processing method can are referred to that in the CEN European Standard (2013). Due to the limitation of test conditions, the wind data are recorded at the height of 3 m from the ground. In order to obtain the wind speed at the bridge height, a power law is used (Frost, 1948; Wu, 2014). Generally, the wind profile of the atmospheric boundary layer (ABL) (from surface to around 200 m) is logarithmic in nature and is best approximated using the log wind profile equation that accounts for surface roughness and atmospheric stability (Cook, 1985). However, sometimes the surface roughness or stability information is not available, so the wind profile power law relationship is often used as a substitute for the log wind profile. In this section, a power law is used to describe the
158
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 1. Layout of measurement points on the noise barrier.
Fig. 2. High-speed railway bridge with noise barriers on both sides: (a) Cross-sectional view, (b) Top view.
Fig. 3. Measurement points on the barriers in tests: (a) Inner surface, (b) Outer surface.
u ¼ ur ðz=zr Þα
vertical distribution of horizontal mean wind speeds, which is: u=ur ¼ ðz=zr Þα
(1)
(2)
where u is the wind speed at height z, and ur is the known wind speed at a reference height zr ¼ 10 m above the ground based on the Chinese Standard, 2012—Load code for the design of building structures (GB50009-2012). The exponent (α) is an empirically derived coefficient
In order to estimate the wind speed at a certain height z, the relationship would be arranged to:
159
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 6. Schematic diagram of data definitions.
and direction in the tests and to obtain the influence of environmental wind on the fluctuating pressure on the noise barriers.
Fig. 4. Arrangement of the velocity radar.
2.3. Data processing To deal with those test data, a sampling frequency 1 kHz and a sixth order Butterworth low-pass filter 40 Hz are utilized. The data definitions diagram is shown in Fig. 6. ΔP indicates the peak-to-peak pressure value with the largest difference between the maximum positive peak (denoted by Pmax), and the maximum negative peak (denoted by Pmin). The results of several runs at the same speed grade were analysed to determine the reliability and repeatability of the test system. The No. 1 measurement point, as shown in Fig. 1, was selected as an example to analyse this system's repeatability, when a 16 cars marshalling CRH380A EMU repeatedly passed it at the speed of 380 km/h. There were three operational runs: as the train ran from north to south, the wind direction was from west to east and the wind speeds were 0.78 m/s, 0.85 m/s and 1.06 m/s. The time history curves of pressure in the 3 runs are illustrated in Fig. 7, and the test results are presented in Table 2. During data processing, if the train speed difference is in the range of approximately 2 km/h to a predetermined speed level, the data is considered to be valid at this level. With respect to the influence of environmental wind on the pressure, when the wind speed is below 2 m/ s, the influence is ignored. At each train speed level, experiments are repeated at least three runs, and then the fluctuating pressure data at each measurement point are obtained by using the ensemble average method. Fig. 7 indicates that the pressure curves of the No. 1 measurement point in 3 runs are basically identical. The pressure changing curves due to the passage of the train head and tail are substantially coincident, and the curves are slightly different when the middle cars pass, due to a small difference in the train speed and environmental wind. Since the environmental wind direction and wind speed are not
Fig. 5. Arrangement of the wind speed sensor.
that varies dependent on the stability of the atmosphere. Because the test point is located at the flat terrain and open land cover conditions, for this terrain, α is approximately 0.15 (GB50009-2012). So in this paper, α ¼ 0.15 is used.
2.2. Test cases As it is known, field tests cost large amounts of resources, leading to limited runs for each train speed, so it is very difficult to obtain lots of data, especially for those in the environmental wind conditions. In this full-scale test, the test runs and corresponding train speeds are listed in Table 1. The train passes the test section on the near rail line 29 runs and on the far rail line 8 runs. Generally, the number of total runs is enough for the field tests, because the environmental wind is always not taken into account, except for the tests in strong wind regions. According to the climate data along the railway in the past years, there are some days with slight wind. Therefore, sensors were set up to monitor the wind speed Table 1 Test number at different train speeds. Train speed (km/h)
Runs Near rail
Far rail
250 300 350 380
4 5 8 12
0 4 4 0
Fig. 7. Time history curves of pressure in the 3 runs. 160
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Table 2 Repeatability of measurement results. Case
Pmax (Pa)
Pmin (Pa)
ΔP (Pa)
Standard deviation (Pa)
Relative standard deviation (%)
The first run The second run The third run
883 903 881
680 650 657
1 563 1 554 1 538
10.34
0.67
However, it still fluctuates in a small limited range around the ambient pressure due to the complex structures of bogies and inter-carriage gaps (CEN European Standard, 2013; Hemida et al., 2014; Huang et al., 2017; Niu et al., 2017; Zhang et al., 2018). When the train tail passes by the measurement point, the pressure first decreases and then increases instantly, and a pressure wave turns up with a negative peak followed by a positive peak. Overall, these typical pressure variations come from the train head and tail, and the pressure caused by the passage of the middle cars is much smaller. The most severe variation of pressure is caused by the passage of the train head, as a result the peak-to-peak values of the head wave are much larger than these of the tail wave, as reported in Zhou et al. (2014), CEN European Standard (2013), Hemida et al. (2014), Huang et al. (2017), Niu et al. (2017), Zhang et al. (2018). Note that the peak-to-peak pressure values on the inner surface of the barrier are much larger than those on the outer surface at the same height, as the air is posed on the points in different impact ways. For the point on the inner surface, the air is pushed by the train directly acts on it, while for the outer one the pressure variation is induced by the air from the railways in an indirect approach. The maximum positive peak and negative peak of the fluctuating pressure on the inner surface of the barrier occurred when the train head passed the measurement point, and the absolute value of the maximum positive peak of the pressure is larger than the maximum negative peak. In contrast, for the tail wave, the absolute value of the negative peak is larger than the positive peak. Therefore, the starting and end points, i.e., the noses, play significant roles in these pressure waves, which contributes to a larger absolute value of pressure. The phenomenon is also found in Baker et al. (2014). In the wake of the high-speed train, the flow is more turbulent and complex, and two counter-rotating vortices are shown (Bell et al., 2016, 2017; Zhang et al., 2016). As a result, more fluctuations from the flow structures are expected to cause the variations of the pressure at the monitoring points after the train travel away, as remarked in red circles in Fig. 8. Additionally, as the previous discussion, this fluctuation of the inner point is much stronger than that of the outer surface, which is also reported by Baker (2014a,b).
repetitive along the railway and the environmental wind speed is less than 2 m/s, the test data in Table 2 present high accuracy with the relative standard deviation of 0.67%. 3. Test results and analysis 3.1. Regularity analysis of the fluctuating pressure on the barrier 3.1.1. Characteristic curve analysis Fig. 8 shows the time history curves of fluctuating pressure on inner and outer surfaces of the noise barrier when a 16 car marshalling CRH380A EMU passes at the speed of 380 km/h. This typical curve has been reported in CEN European Standard, 2013 and Zhou et al. (2014). The fluctuating pressure is generated when the train passes close to the noise barrier, which is a dynamic process changing over time. It is interesting to note that this history curve is similar to that the pressure sweeps along a line that is parallel to the length of the train with the same distance (CEN European Standard, 2013; Hemida et al., 2014; Huang et al., 2017; Niu et al., 2017; Zhang et al., 2018). When the train head approaches the measurement point, the pressure starts to climb up to the highest. This is because the train pushes air towards the inner surface of the barrier. The positive peak occurs at the arrival of the train nose. After that, the pressure drops rapidly to a negative peak. Therefore, a complete pressure wave is generated by the train head. After the train head passes the measurement point, as it is found in (CEN European Standard, 2013; Zhou et al., 2014), the pressure fluctuation at the observation point considerably slows down, as the cross section of the train is uniform.
3.1.2. Relationship curves between the fluctuating pressure and train speed Fig. 9 shows the fluctuating pressure curves of the No.1 measurement point (Inner surface) and the No. 8 measurement point (Outer surface) on the barrier, when a 16 car marshalling CRH380A EMU passes close to the noise barrier at different speeds. Here to show the differences of fluctuating pressure curves at different train speeds, it is supposed that the tail nose of the train leaves the points at the same time, because in this curve the most severe variation of pressure comes from the train head passing. Clearly a higher train speed contributes to a larger peak-to-peak pressure of the head wave no matter where the monitoring points are. Meanwhile, the fluctuations induced by the middle cars also enhance with the train speed, seeing Fig. 9. Fig. 10 shows the law curves of ΔP, defined as the difference of Pmax to Pmin in Fig. 6, on the inner and outer surfaces of the barrier with changes of the train speed. According to the research conducted by Tokunaga et al. (2016), Tian (2007) and Zhou et al. (2014), the ΔP values of the head wave are approximately proportional to the square of the train speed. Here, this law is clearly shown on the inner surface of the barrier. The pressure amplitudes on the outer surface at different locations also show power laws with respect to the train speed. However, these exponents that present some differences, as shown in Fig. 10 (b), reduce with decrease in the height of monitoring points. Compared to the pressure
Fig. 8. Time history curves of fluctuating pressure on measurement points: (a) Inner surface, (b) Outer surface. 161
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 9. Fluctuating pressure curves of different measurement points: (a) Point 1 (b) Point 8.
Fig. 10. Relationship curves between ΔP and train speeds: (a) Inner surface, (b) Outer surface.
amplitudes on the inner surface of the barrier, seeing Fig. 10(a), these amplitudes are so small that may be easily affected by the measurement accuracy of the sensors and environmental wind. Therefore, roughly, the pressure amplitudes are proportional to the square of the train speed no matter on the inner and outer surfaces of the barrier. To analyse the train speed impact on the fluctuating pressure on barrier surfaces, a peak-to-peak time interval, denoted by Δt, is the time from Pmax to Pmin of the head wave is defined according to CEN European Standard (2013). Where, Ln is the length of the train head and vtr is the train speed. Table 3 shows the Δt, when a 16 grouping-car CRH380A EMU passes the No.1 measurement point at different train speeds. Δt Ln = vtr
Table 3 Peak-to-peak time intervals. Train running speed(km/h)
250
300
350
380
Δt (s) Ln (m) from tests Ln (m) from a realistic train
0.1099 7.6319 12.0000
0.0917 7.6389
0.0786 7.6384
0.0723 7.6317
distance from the far railway line is 8.40 m ΔPnear is defined as the ΔP at the points when the train runs on the near railway line, while ΔPfar is the ΔP at the points when the train runs on the far railway line. The general observation is that when the lateral distance between the train and the barrier is smaller, the train induced pressure on the point fluctuates largely, especially for the head wave, as shown in Fig. 12(a). This is because the energy of the wave is diffused in the propagation. It is similar to these shown in Fig. 12(b) that the ΔP reduces with the increase in height of monitoring points. However, due to the same scale for the ΔPnear and ΔPfar, it is a little difficult to see the decrease trend for ΔPfar. Overall, the ratios ofΔPnear to ΔPfar for these 6 monitoring points are in a range of 0.245–0.262.
(3)
As the train speed increases, the peak-to-peak time interval Δt decreases gradually and the corresponding slope becomes increasingly steep. According to formula (3), the train head length is obtained from Ln Δt * vtr, as shown in Fig. 11(a). Note that this length actually corresponds to the length from the front tip of the nose to the transition curve position of the cabin window instead of the 12 m long train head of a realistic high-speed train, seeing Fig. 11(b). Furthermore, the position of the maximum negative pressure on the train body surface in the test is essentially consistent with that previously reported by Yang et al. (2015) when a high-speed train passes on the bridge. The result shows that there is a debate about the definition of the length Ln of the train head in aerodynamics, and the parameters require to be redefined.
3.1.4. Comparison with other sources of data From Section 3.1.1, it is shown that the absolute value of the positive peak of the fluctuating pressure is larger than that of the negative peak. So in this section, the positive head pulse pressure coefficients of the test results are compared with other experimental data and some data from empirical equations, as shown in Fig. 13. The measurements were conducted by Tokunaga et al. (2016) and Horie and Sugiyama (1986). For the equations, one is included in the 2013 EN (CEN European Standard, 2013) that is Cp ¼ 2.5/(Yþ0.25)2þ0.02, where the Y is the distance from the track centre, which is greater than 2.3 m. The other equation is from a literature published by Baker et al. (2014), as follows: Cp ¼ 6/(Yþ1.75)2. When the distance between the noise barrier and the track centre is
3.1.3. Comparative analysis of the fluctuating pressure and the train running lines Fig. 12 (a) shows the time history curves of fluctuating pressure at the No. 1 measurement point on the inner surface of the noise barrier when a train runs on the near and far rails at the speed of 350 km/h, respectively. A comparison of ΔP at points No. 1–6 is found in Fig. 12(b). The distance of the inner surface of the barrier from the centreline of the track is 3.40 m when the CRH380A EMU runs on the near railway line, and the 162
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 11. Length of the train head: (a) Test length, (b) Actual length.
pressures and forces on noise barrier are underestimated. Therefore, it seems better to adopt the equation from Baker et al. (2014) to estimate the transient aerodynamic pressures on the noise barrier. Additionally, the experimental results of the transient aerodynamic pressure coefficients in Japan are less than the calculated data using the equation in the 2013 EN (CEN European Standard, 2013). Based on above analysis, although the empirical equations can be used to estimate the transient aerodynamic pressures on the noise barriers in China, these data still show slight differences compared with the data in the present work, which is not in accordance with the actual situation of China high-speed railways. Therefore, it is very important to obtain those data directly coming from China high-speed railways. 3.2. Distribution law of ΔP along the vertical direction on the noise barrier Fig. 14 shows the changing trends of ΔP on the inner and outer surfaces of the noise barrier along the height direction when a 16 car marshalling CRH380A EMU passes the test points on the noise barrier at different speeds. As illustrated in Fig. 14, the ΔP values on the inner surface of the noise barrier decrease with the increase in the height of measurement points from bottom to top. At the highest point they are approximately 74.5%– 81.9% of these at the bottom when the CRH380A EMU passes the test section at different speeds. The ΔP distributions on the noise barrier shows a slight change from the centre to the bottom, and the values at the centre are approximately 92.4%–95.2% of these at the bottom. The phenomenon is also found in the report (PECþS, 2007) and the height of the noise barrier is 4.0 m. This is mainly because the effective airflow area at the bottom of the noise barrier is less than that at the top, and the airflow velocity at the bottom is higher than that at the top. Furthermore, the direction of airflow shows an outflow trend to the outer surface of the noise barrier, and thus the pressure at the bottom of the noise barriers is greater than that at the top. The ΔP values at the bottom of the outer surface of the noise barrier are the smallest along the height direction. Then they gradually increase with the height. However, at the bottom of the inner surface of the noise barriers they are the largest. Due to the shielding effect of noise barriers, the fluctuating pressure generates when the train passes the measurement points and reaches at the top of the outer surface of the noise barriers, then spread to the bottom of the outer surface down along the noise barriers. This result in the ΔP values of the measurement points at the top the barriers exceeding that at the bottom. As the ΔP values on the inner surface close to the bottom of noise barriers are the largest, it is required to design a more appropriate bottom of the noise barrier.
Fig. 12. Pressure change curves and peak-to-peak comparison: (a) Pressure variation at point 1, (b) Peak-to-peak comparison of fluctuating pressure at points 1-6.
3.3. Fluctuating pressure distribution law on the noise barriers along the train running direction
Fig. 13. Comparison of pressure coefficients with previous work.
3.40 m, the full scale pressure coefficients on the inner surface at different train speeds are slightly larger than those calculated data using the recommended equation in the 2013 EN (CEN European Standard, 2013), whereas these coefficients are smaller than those calculated data using the proposed equation from Baker et al. (2014). If the equation in the 2013 EN (CEN European Standard, 2013) is used to calculate the pressure coefficient of the head pulse, the transient aerodynamic
Fig. 15 shows the changing curves and ΔP comparison chart at different measurement points which are fixed on the inner surface of the noise barrier at the same height but different longitudinal positions, when the 16 car marshalling CRH380A EMU passes the test points at different speeds. Although those points are at the same height, there are some distances in the train running direction among those points, especially for the point 16 with a distance of 150 m from point 4 which is very 163
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Fig. 16. Time history curves of fluctuating pressure and peak-to-peak comparison at the measurement points on the inner surface: (a) Time history curves of fluctuating pressure, (b) Peak-to-peak comparison chart of fluctuating pressure.
Fig. 14. Changing trends of peak-to-peak pressure values along the height direction: (a) Inner surface, (b) Outer surface.
close to points 12 and 15, as shown in Fig. 1. As a result, the time when the train head arrives those points is different, as shown in Fig. 15(a). Since the heights are the same for all points, according to CEN European Standard (2013), the time history curves of all points should be very similar. In fact, from the measurement, the ΔP values of monitoring points along the train running direction are almost the same when the points are located in the middle of the noise barrier. Additionally, when a 16 car marshalling CRH380A EMU passes the test points, the ratios of the maximum to the minimum of the peak-to-peak pressure values at different points with the same height are between 101.0% and 101.7%. This error comes from the measurement error of the test system and the relative height between the bottom of the barrier and the rail top. 3.4. Influence of different train marshalling lengths on the fluctuating pressure Fig. 16 shows the time history curves of fluctuating pressure, Fig. 16(a), and the ΔP values comparison chart, Fig. 16(b), when a CRH380A EMU with 2 types of marshalling (16 car-grouping and 8 cargrouping) passes close to the noise barrier at 380 km/h and the environmental wind speed is less than 2 m/s. The previous analysis indicates that the fluctuating pressure distribution on the surface of the noise barriers mainly depends on the train speed, the streamlined head and the couplings (CEN European Standard, 2013; Baker, 2014a,b). In the present work, the streamlined head and the train speed are the same and there is no coupling in the 16 car-grouping CRH380A EMU, so the ΔP values at the identical point should be basically the same. In fact, when the CRH380A EMU with the 16 car and 8 car marshalling types passes the measurement points on the noise barrier at the same speed, the head wave and the ΔP values at the same measurement point are basically the same, and the largest difference between
Fig. 15. Time history curves of fluctuating pressure and peak-to-peak comparison chart at different measurement points with the same height: (a) Time history curves of fluctuating pressure, (b) Peak-to-peak comparison chart of fluctuating pressure.
164
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
Table 4 Influence of environmental wind on the peak-to-peak pressure values. Running direction
Environmental wind speed (m/s)
Environmental wind direction
ΔP (Pa)
From From From From From
8.06 5.86 1.44 5.18 6.53
Due north Due north Due north Due north Due north
1 383 1 434 1 599 1 679 1764
north to south north to south south to north south to north south to north
them is only 1.74%, which further confirms the fluctuating pressure distribution on the surface of the noise barriers and the ΔP mainly depends on the train speed, the streamlined head and the couplings.
pressure values is 27.6% larger than the minimum, when the environmental wind speed changes within the range of 0–8.06 m/s, and the 16 car marshalling CRH380A EMU runs on the line close to the noise barrier in different directions at the speed of 380 km/h. Given that the environmental wind velocity is generally small in the tests and the direction is relatively fixed, this study only analyses the impact of the environmental wind velocity on the fluctuating pressure. Hence, an in-depth analysis of the environmental wind direction on the influence of fluctuating pressure is not performed, which are continued to study by numerical simulations in the next work.
3.5. Influence of environmental wind on the fluctuating pressure on noise barriers In the tests, 29 runs on the near rail line and 8 runs on the far rail line were carried out, respectively, as listed in Table 1. The number of total runs is sufficient for a conventional field test, and 5 runs come across environmental wind if the wind speeds recorded by the wind speed and direction sensors as shown in Fig. 5 are not less than 1 m/s. Although the samples seem not enough to be used for concluding the influence laws of environmental wind on the fluctuating pressure on noise barriers, general relationships between the wind velocity, wind direction and the peak-to-peak pressure measured are obtained. Also those data are used to validate against the numerical simulations in the next work. The ΔP values at the measurement point 1 as a 16 car marshalling CRH380A EMU passes close to the noise barrier at the speed of 380 km/h under different environmental wind speeds and directions as shown in Table 4. Fig. 17 shows the time history curves of fluctuating pressure at point 1 on the inner surface of the noise barrier under different wind environments, when the 16 car marshalling CRH380A EMU passes the point at the speed of 380 km/h. The peak-to-peak pressure values on the noise barrier surfaces are significantly affected by the environmental wind. When the CRH380A EMU runs against the wind direction, the peakto-peak pressure values at the measurement point on the inner surface of the noise barrier are larger than these when the train runs along the wind direction. This is because the relative airflow speed at the measurement points on the noise barrier surfaces are superimposed by both the flow velocity induced by the CRH380A EMU and wind velocity component along the line direction in the existence of environmental wind. In other words, when the train runs against the wind, the relative wind speed at the measurement point is the sum of the flow velocity induced by the CRH380A EMU and wind velocity component along the line direction. In contrast, the relative wind speed is the difference between the flow velocity induced by the CRH380A EMU and wind velocity component along the line direction. Moreover, the maximum in these peak-to-peak
4. Conclusions A full scale field test has been conducted to obtain the variations of pressure on a 2.15 m high bridge noise barrier induced by the passage of a CRH380A EMU high-speed train at the speed of 250 km/h ~380 km/h using LL-250 type pressure sensors which are produced by the Kulite Company in USA. A Stalker Speed Sensor (S3) Police Option radar with high accuracy is used to measure the train speed when the train passes by the testing region, and XFY3 type wind speed and direction sensors are installed in the upstream of the bridge to check the wind speeds and directions. The time history curves of fluctuating pressure on the inner and outer surfaces of the barrier, especially the peak-to-peak pressures caused by the passage of the train head and tail, as found by previous research in scaled, numerical and full scale studies. The positive head pulse pressure coefficients of the test results show slight differences, compared with other experimental data and some data from empirical equations. Further, train noses have significant effects on the first part of the head and tail pressure waves, i.e., a larger absolute value for the first peak in both waves. The ΔP distributions on the inner and outer surfaces of the barrier are not identical; indicating the pressure on the outer surfaces is induced by the air from the railways in an indirect approach and is easily affected by the ambient environment. Due to two counterrotating vortices in the wake of the train tail, the pressure on the barrier continues fluctuating after the train leaves. The ΔP values of the head wave are approximately proportional to the square of the train speed, as reported in previous investigations. With the increase of the train speed, the peak-to-peak time intervals of ΔP decrease gradually and the slopes become increasingly steep. The findings from the tests indicate that for a CRH380A EMU the aerodynamic length of the train head is between 7.63 and 7.64 m. This is different from the physical length of the realistic train head with 12.00 m. Thus, there is a discrepancy with respect to the definition of the train head length (denoted by Ln), and the parameters need to be redefined. The distance of the barrier from the train symmetry plane determines the ΔP value. When the lateral distance between the train and the barrier is smaller, the train induced pressure on the point fluctuates largely, especially for the head wave. Generally, the ratios of ΔPnear to ΔPfar are in a range of 0.245–0.262. The train marshalling length does show much effect on the fluctuating pressure time history curve and ΔP, if there are no couplings in a long marshalling train. Along the noise barrier height, the ΔP values on the inner surface of the noise barrier decrease with the increase in the height of measurement points. Additionally, the ΔP values on the outer surface of the noise barrier increase with the height increasing. Taking the wind speed and direction into account, when the environmental wind speed changes
Fig. 17. Curves of fluctuating pressure at point 1 on the inner surface of the noise barrier under different wind environments. 165
X.-h. Xiong et al.
Journal of Wind Engineering & Industrial Aerodynamics 177 (2018) 157–166
within the range of 0–8.06 m/s and the 16 car marshalling CRH380A EMU runs on the line close to the noise barrier in different directions at the speed of 380 km/h, the maximum in these ΔP values is 27.6% larger than the minimum. Overall, the results in this paper can provide data support for the study on the dynamic response evaluation of high-speed railway noise barriers under the fluctuating loads and also contribute to the data reference for revising and improving the relevant standards for the noise barriers along high-speed railways.
Chen, X.D., Li, S.D., Wang, Z.H., 2011. Numerical simulation study on high-speed train induced impulsive pressure on railway noise barrier based on ALE. J. China Railw. Soc. 33 (12), 21–26 (in Chinese). Chinese Standard, 2012. Load Code for the Design of Building Structures. GB50009—2012. Cook, N.J., 1985. The Designer's Guide to Wind Loading of Building Structures: Part 1. Butterworths. Frost, R., 1948. Atmospheric turbulence. Q. J. R. Meteorol. Soc. 74 (321–322), 316–338. Hemida, H., Baker, C.J., Gao, G., 2014. The calculation of train slipstreams using largeeddy simulation. Proc. Inst. Mech. Eng. - Part F J. Rail Rapid Transit 228 (1), 25–36. Holmes, S., Schroeder, M., 2002. Aerodynamic Effects of High-speed Passenger Trains on Other Trains. DOT/FRA/ORD-01/12. Horie, A., Sugiyama, T., 1986. Field Test of Train Draft on Tohoku Shinkansen. Technical Report A-86,Railway Technical Research pre-Report. Huang, S., Hemida, H., Yang, M.Z., 2017. Numerical calculation of the slipstream generated by a CRH2 high-speed train. Proc. Inst. Mech. Eng., Part F: J. Rail. Rap. Tran 230, 103–116. Lee, H., 1999. Assessment of Potential Aerodynamic Effect on Personnel and Equipment in Proximity to High-speed Trains Operations. DOT/FRA/ORD-99/11. Long, L.P., Zhao, L.B., Liu, L.D., 2010. Research on the air turbulent force loading on noise barrier caused by train. Eng. Mech. 27 (3), 246–250 (in Chinese). Chinese National Railway Standard, 2014. Code for Design of High Speed Railway, TB 10621. Niu, J.Q., Zhou, D., Liang, X.F., 2017. Numerical simulation of the effects of obstacle deflectors on the aerodynamic performance of stationary high-speed trains at two yaw angles. Proc. Inst. Mech. Eng. - Part F J. Rail Rapid Transit 232. Schetz, J.A., 2001. Aerodynamics of high-speed trains. Annu. Rev. Fluid Mech. 53 (2), 371–414. Tian, H.Q., 2007. Train Aerodynamics. China Railway Publishing House (In Chinese). Tokunaga, M., Sogabe, M., Santo, T., Ono, K., 2016. Dynamic response evaluation of tall noise barrier on high speed railway structures. J. Sound Vib. 366, 293–308. Wu, Y., 2014. Wind Engineering and Structure Wind-resistant Design. Harbin Institute of Technology Publishing House (In Chinese). Yang, N., Zheng, X.K., Zhang, J., Law, S.S., Yang, Q.S., 2015. Experimental and numerical studies on aerodynamic loads on an overhead bridge due to passage of high-speed train. J. Wind Eng. Ind. Aerod. 140 (10), 19–33. Zhang, J.W., Lü, J.P., Tu, Y.M., Kong, C.W., Ge, W.J., 2009. Numerical analysis of highspeed train induced impulsive pressure on railway bridge's noise barrier and its flange Plate with collision-preventing wall. China Railw. Sci. 30 (3), 27–32 (In Chinese). Zhang, J., Li, J.J., Tian, H.Q., Gao, G.J., Sheridan, J., 2016. Impact of ground and wheel boundary conditions on numerical simulation of the high-speed train aerodynamic performance. J. Fluid Struct. 61, 249–261. Zhang, J., Wang, J.B., Wang, Q.X., Xiong, X.H., Gao, G.J., 2018. A study of the influence of bogie cut outs' angles on the aerodynamic performance of a high-speed train. J. Wind Eng. Ind. Aerod. 175, 153–168. Zhou, D., Tian, H.Q., Zhang, J., Yang, M.Z., 2014. Pressure transients induced by a high speed train passing through a station. J. Wind Eng. Ind. Aerod. 135, 1–9. Zhu, Z.Q., Cheng, Z.Q., 2011. Numerical simulation and experimental verification of dynamic pressure on noise barriers of high speed railways. Railway standard design. 11, 77–80 (in Chinese).
Acknowledgments This work was supported by thanks the National key R & D program of China (Grant No. 2016YFB1200506-03), the National Science Foundation of China (Grant Nos. 51205418, 51605044 and U1334205), the Science Foundation of Hunan Province of China (Grant No. 2016jj3004) and the Science and Technology Research Programme of China Railway Corporation (Grant No.2016T004-C). References Baker, C.J., Jordan, S., Gilbert, T., Quinn, A., Sterling, M., Johnson, T., Lane, J., 2014. Transient aerodynamic pressures and forces on trackside and overhead structures due to passing trains. Part1 Model scale experiments. Part2 Standards applications. Proc IMechE Part F: J.Rail Rapid Transit 228 (1), 37–70. Planning Engineering Consulting þ Services China Ltd, 2007. Consultation Report of the Noise Barriers in Chinese Railway Passenger Dedicated Line. Baker, C.J., 2010. The flow around high speed trains. J. Wind Eng. Ind. Aerod. 98, 277–298. Baker, C.J., 2014a. A review of train aerodynamics part1-fundamentals. Aeronaut. J. 118 (1201), 201–228. Baker, C.J., 2014b. A review of train aerodynamics part1-applications. Aeronaut. J. 118 (1201), 345–382. Bell, J.R., Burton, D., Thompson, M.C., Herbst, A.H., Sheridan, J., 2016. Flow topology and unsteady features of the wake of a generic high-speed train. J. Fluid Struct. 61, 168–183. Bell, J.R., Burton, D., Thompson, M.C., Herbst, A.H., Sheridan, J., 2017. A wind-tunnel methodology for assessing the slipstream of high-speed trains. J. Wind Eng. Ind. Aerod. 166, 1–19. Carassale, L., Brunenghi, M.M., 2013. Dynamic response of track side structures due to the aerodynamic effects produced by passing trains. J. Wind Eng. Ind. Aerod. 123, 317–324. CEN European Standard, 2013. Railway applications – aerodynamics. In: Part4: Requirements and Test Procedure for Aerodynamics on Open Track. CEN EN 14067–4.
166