Analysis of the dynamic response of the new interchange channel closely under the existing line Authors: Yunsi Liu, Yueyuan Huang, Yu Huang, Yonglin An, Yaxiong Peng Published: 2024 ABSTRACT AbstractThe vibration caused by the existing operating metro has a negative impact on the construction of new metro lines and the nearby surroundings. The focus of this study is the confined interchange area between Beijing Metro Line 17 and Shilihe Station of Line 10. In the study area, the interchange between Shilihe Station of Line 10 and Line 17 has a confined frame structure that is already in operation without any overburden between the existing line and the new line; therefore, the vibration of the operating subway interferes with the construction. The measured and numerical simulation results show that the vibration response has an attenuating effect in the frame structure from top to bottom, and these studies are related to the train operating speeds of 20 km/h, 40 km/h, 60 km/h and 80 km/h. The type and spacing of the I-beams also have an effect, and the higher the train speed, the greater the vibration response. The principle of equivalent stiffness of I-beam is proposed, and the functional relationship between equivalent stiffness and I-beam parameters is deduced. The models of no vibration amplification effect, weak vibration amplification effect and strong vibration amplification effect are established, and it is obtained that when the I-beam is I20b and the spacing is 0.8 m, the equivalent stiffness is maximum, the vibration response of the frame structure is minimum, the vibration load is attenuated in the frame structure, and the vibration amplification effect of the soil on both sides is smaller. The study of the vibration amplification effect provides some theoretical guidance for the construction of similar projects.Keywords: Train vibration, Flat-topped and straight-walled station, Close-fitting, Frame structure, I-beamSubject terms: Engineering, Civil engineering Introduction As subway construction progresses, numerous challenges have arisen in the construction of tunnels employing the concealed excavation method1,2. For example, the construction of new metro lines cannot avoid crossing existing lines3,4Simultaneously, the vibrations emanating from train operations on the existing line will exert an impact on the construction of the adjacent, newly excavated line5–8Studying the dynamic response characteristics and displacement change patterns of newly constructed closely spaced subway tunnels, subject to vibration loads from overhead trains, is of paramount importance for ensuring the long-term safety and stability of tunnel operations9–11. The researchers developed a track-tunnel-ground interaction model to determine the propagation characteristics of train vibrations. By analysing acceleration and velocity. The results showed that a vibration amplification zone exists at a certain distance from the vibration centre and that structures located in the vibration amplification zone will be sensitive to ground vibrations caused by metro trains12–14. A method has been proposed that can effectively predict tunnel and ground vibrations caused by metro trains. Firstly, the vehicle-track subsystem is used to determine the track-tunnel interaction forces, and then the resulting forces are applied to the 2.5D FEM-PML tunnel-soil system, and eventually the accuracy is verified by comparison15. A numerical track-tunnel-strata model has been developed to study the dynamic response of tunnels under train vibration loads, and the vibration response law of metro operating trains in a composite layer of clay and silt has been analysed, and the long-term cumulative deformation of the powder-fine sand layer under train loads has been investigated using a layered summation method16,17.The measured vibration waves are often accompanied by noise and other influencing factors, which need to be filtered. To overcome the limitation that the variational mode decomposition algorithm cannot flexibly adjust the frequency range and spectral location to decompose the signal, has been proposed to decompose the measured signal18,19. A modified multiplicative alternating direction method is used to solve the formed variational model, thus realising the multi-channel and fixed frequency decomposition of vibration waves. Increasing the material strength of the frame structure contributes to the overall stability, with lower modulus of elasticity frame structure materials having lower strength and higher losses. As a result, vibrations in sensitive areas are reduced while avoiding wave reflection into the track20,21. Underpass tunnel excavation inevitably causes superstructure movements and a series of numerical analyses have been carried out to investigate the behaviour and damage mechanisms of frame buildings with independent foundations due to tunnel-induced ground movements22,23. However, less research has been carried out on the analysis of the dynamic response and support effect under the influence of train vibration loads for a metro station with a flat roof and straight walls. Therefore, this study first analyses the dynamic response law through field tests, and then builds a 3D numerical model through Flac3D to analyse the variation law of the displacement of the close-fitting part caused by the arrangement of different types of I-beams and different spacing of I-beams under the action of train vibration load, which provides a new idea for the construction of other similar close-fitting stations under the action of train vibration load. It provides a new idea for the construction of other similar close-packed stations under train vibration loads. On-site measurement Engineering background This study focuses on the vibration generated by the existing tunnel operation train on the new subway tunnel structure vibration response and deformation analysis, relying on the project to build the new Beijing Metro Line 17 Shilihe Station for the three lines of the interchange, and the upper part of the existing subway Line 10 and Line 14 Shilihe Station to form a “T”-type interchange, as shown in Fig. 1. Diagram of the close paste part. The new subway Line 17 Shilihe Station interchange area is closely connected to the existing subway Line 10 Shilihe Station, with a length of 21.6 m, a width of 16.05 m, and a height of 8.32 m.The vibration generated by the train operation will not only pose vibration and noise hazards to existing structures and the surrounding residential environment, but also cause significant deformation and potential damage, particularly to the neighboring subway lines under construction.When the existing Line 10 subway line is in operation, the vibration generated by the train is propagated through the bottom plate of the existing station to the top, left and right sidewalls, bottom, and excavation surface of the new interchange structure. This results in spalling of the shotcrete in the new interchange’s initial support, weakening its overall stability, and causing deformation of the surrounding rock that exceeds the allowable deformation limit., as shown in Fig. 2. Metro line 17 interchange excavation site photos. Combined with the borehole ground investigation data, the soil layer around the Shilihe Station of the new Metro Line 17 can be categorized into four layers, which are closely connected with the existing Shilihe Station of Metro Line 10: (1) he first layer consists of vegetative fill, with an average thickness of 15 m, positioned above the existing Metro Line (2) The second layer comprises pulverized clay, with an average thickness of approximately 15 m. The existing Metro Line 10 is situated within this layer.(3)The third layer consists primarily of fine sand, with an average thickness of 10 m, and it is within this layer that the new Metro Line 17 is situated.(4) The fourth layer is primarily composed of silty clay, situated beneath the newly constructed Metro Line 17,Each soil layer is shown in Fig. 3. Schematic diagram of soil types and distribution. The operating train of Beijing Metro Line 10 is a 6-section A-type subway vehicle, and the characteristic length and basic power parameters of the vehicle are shown in Fig. 4; Table 1, respectively. Table 1Basic dynamic parameters.ParametersVehicleUnitsCar body mass (Mc)35. 0tBogie frame mass(Mt)5. 5tWheelset mass(Mw)2. 2tMoment of inertia of car body(Jc)1712. 6t∙m2Moment of inertia of bogie frame (Jt)5. 1t∙m2Primary suspension stiffness (per axle box)(Kp)1. 68MN/mPrimary suspension damping (per axle box)(Cp)29. 0kNs/mSecondary suspension stiffness (per bogie) (Ks)0. 32MN/mSecondary suspension stiffness (per bogie) (Cs)17. 2kNs/m Basic dynamic parameters. Train size characteristics. The transfer tunnel for the new Metro Line 17 adopts the flat top straight wall CRD method in combination with the self-elevation construction method. The primary support system employs spray anchor reinforcement combined with I18 steel beams. The excavation sequence for the pilot holes is to excavate one side of pilot holes 1 and 2 first. The excavation support for the upper pilot hole 1 is completed after the excavation of the lower pilot hole 2. The surrounding rock is then stabilized, followed by the excavation of the other side of pilot holes 3 and 4, as shown in Fig. 5. CRD method for flat-topped straight walls. Analysis of real train vibration Vibration signal measurement In order to analyze the impact of the vibration of the train operating on the lower part of the new frame structure, this study adopts Sichuan Jiaobo L20-S vibration tester to carry out the monitoring of the acceleration generated by the train vibration on the structure.L20-S vibration tester trigger mode is level trigger, trigger level is 0. 001 m/s2, storage time is 20s. Three groups of monitoring points are arranged in the existing Metro Line 10 stations, namely the left side of the station (P1), the middle of the station (P2) and the right side of the station (P3), as shown in Fig. 6.Five radial (X), tangential (Y) and vertical (Z) vibration acceleration signals are recorded at each group of monitoring points when different trains pass through, and the recording results are shown in Table 2.From the results in Table 2, it can be seen that the acceleration generated by the vibration of the Metro Line 10 train is mainly distributed in the range of 0.28 ~ 0.35 m/s², indicating that the acceleration levels of the vibration sources are approximately the same.The results of this study suggest that the vibration responses induced by different trains can be put together for comparative analysis in subsequent vibration analyses. There are monitoring points inside the subway station of Line 10. Table 2Vibration acceleration at the monitoring point in the station of Metro Line 10 is available.MeasurementnumberAcceleration( m/s2)MeasurementnumberAcceleration( m/s2)MeasurementnumberAcceleration( m/s2)XYZXYZXYZP1—#C10.0320. 0280. 030P2—#C60. 0260.0270. 025P1—#C110. 0300.0280.029P1—#C20.0280. 0240. 026P2—#C70. 0270.0250. 024P1—#C120. 0310.0300.030P1—#C30.0290. 0260. 028P2—#C80. 0300.0290. 031P1—#C130. 0280.0260.024P1—#C40.0240. 0180. 022P2—#C90. 0320.0300. 028P1—#C140. 0240.0220.023P1—#C50.0310. 0190. 030P2—#C100. 0280.0240. 027P1—#C150. 0270.0280.025 Vibration acceleration at the monitoring point in the station of Metro Line 10 is available. From the results depicted in Fig. 7, it is evident that the characteristic frequency distributions of horizontal acceleration for the three groups of monitoring points lie within the ranges of 10–70 Hz, 110–190 Hz, and 280–400 Hz. Additionally, the characteristic frequency distributions of vertical vibration acceleration are found to be within the ranges of 50–120 Hz and 170–300 Hz. The relationship between horizontal and vertical acceleration and frequency.(a) P1 monitoring point Frequency characteristics. (b) P2 monitoring point Frequency characteristics. (c) P3 monitoring point Frequency characteristics. To analyze the vibration response characteristics of the interchange tunnel for the new Subway Line 17 where it crosses and overlaps with the existing Subway Line 10, three monitoring sections, labeled ①, ②, and ③, have been established. Each section is spaced 4 m apart, as illustrated in Fig. 8a. Within each monitoring section, eight points are arranged: a1, a2, a3, a4, b1, b2, b3, and b4, as shown in Fig. 8b . Schematic layout of monitoring points. (a) Schematic diagram of monitoring section (b)Schematic diagram of the monitoring points of the frame structure. Interchange tunnel measurement results and analysis Fig. 9 shows the vertical and horizontal acceleration as a function of time for each monitoring point.By analyzing the results of each monitoring point, it can be seen that when the train passes through the wireline, the time domain waveforms of the upper guideway monitoring points a1,a2,a3 will have an obvious peak, and then the dynamic response of the acceleration tends to be stable, and the acceleration response time is maintained at about 6s.The peak acceleration of vibration at each monitoring point can be expressed in terms of Peak Point Acceleration (PPA). The relationship between the horizontal and vertical acceleration and the time of the monitoring points.(a) Vibration Propagation Schematic. (b) ~(h) sequentially represent the measurement values of monitoring points a1 ~ a4 and b2 ~ b4 on the cross-section. Figure 10shows a plot of vertical and horizontal point peak acceleration (PPA) versus each monitoring point on each section: The PPA measurement values of the monitoring points in each monitoring section.(a)Vertical acceleration Comparison of measured vibration acceleration. (b)Horizontal acceleration comparison of measured vibration acceleration. During the downward propagation of the train vibration wave from the bottom plate of the existing line, vibration attenuation is evident in the upper guide hole, and the vibration tends to stabilize as it reaches the lower guide hole. For example, regarding vertical acceleration, the maximum value for the top plate (a1) of the upper guide hole is 0.0113 m/s², while the middle partition (a3) exhibits a maximum vertical acceleration of 0.0026 m/s², showing a 77.0% attenuation. The bottom plate (b3) of the lower guide hole has a maximum vertical acceleration of 0.0006 m/s², demonstrating a 94.7% attenuation, as shown in Fig. 10a. The maximum horizontal acceleration of the top plate (a1) of the upper guide hole is 0.0142 m/s², while the maximum horizontal acceleration of the middle partition (a3) is 0.0038 m/s², representing a 73.2% attenuation. Similarly, the maximum horizontal acceleration of the bottom plate (b3) of the lower guide hole is 0.0008 m/s², which indicates a 94.4% attenuation, as shown in Fig. 10(b). During the downward propagation of the train vibration wave from the bottom plate of the existing line, vibration attenuation is evident in the upper guide hole, and the vibration tends to stabilize as it reaches the lower guide hole. For example, regarding vertical acceleration, the maximum value for the top plate (a1) of the upper guide hole is 0.0113 m/s², while the middle partition (a3) exhibits a maximum vertical acceleration of 0.0026 m/s², showing a 77.0% attenuation. The bottom plate (b3) of the lower guide hole has a maximum vertical acceleration of 0.0006 m/s², demonstrating a 94.7% attenuation, as shown in Fig. 10a. The maximum horizontal acceleration of the top plate (a1) of the upper guide hole is 0.0142 m/s², while the maximum horizontal acceleration of the middle partition (a3) is 0.0038 m/s², representing a 73.2% attenuation. Similarly, the maximum horizontal acceleration of the bottom plate (b3) of the lower guide hole is 0.0008 m/s², which indicates a 94.4% attenuation, as shown in Fig. 10(b). The greater the horizontal distance from the train vibration load, the smaller the corresponding acceleration becomes. For example, along the excavation direction, in section ①, the maximum horizontal acceleration at the top of the upper guide hole (a1) is 0.0142 m/s²; in section ②, it is 0.0096 m/s², with an attenuation rate of 32.4%; and in section ③, it is 0.0068 m/s², with an attenuation rate of 52.1%. Similarly, for the maximum vertical acceleration at the top of the upper guide hole (a1), in section ① it is 0.0113 m/s²; in section ②, it is 0.0079 m/s², with an attenuation rate of 30.1%; and in section ③, it is 0.0045 m/s², with an attenuation rate of 60.18%. The greater the horizontal distance from the train vibration load, the smaller the corresponding acceleration becomes. For example, along the excavation direction, in section ①, the maximum horizontal acceleration at the top of the upper guide hole (a1) is 0.0142 m/s²; in section ②, it is 0.0096 m/s², with an attenuation rate of 32.4%; and in section ③, it is 0.0068 m/s², with an attenuation rate of 52.1%. Similarly, for the maximum vertical acceleration at the top of the upper guide hole (a1), in section ① it is 0.0113 m/s²; in section ②, it is 0.0079 m/s², with an attenuation rate of 30.1%; and in section ③, it is 0.0045 m/s², with an attenuation rate of 60.18%. The horizontal vibration accelerations of the two side walls of the guide tunnel on the interchange channel differ, and a vibration amplification effect is observed. For instance, the vibration acceleration at point a2 in section ①, adjacent to the surrounding rock as shown in Fig. 10b, is 0.0082 m/s², whereas the vibration acceleration at point a4 is 0.0038 m/s². The difference between the two accelerations is 0.0044 m/s². The horizontal vibration accelerations of the two side walls of the guide tunnel on the interchange channel differ, and a vibration amplification effect is observed. For instance, the vibration acceleration at point a2 in section ①, adjacent to the surrounding rock as shown in Fig. 10b, is 0.0082 m/s², whereas the vibration acceleration at point a4 is 0.0038 m/s². The difference between the two accelerations is 0.0044 m/s². Numerical simulation Model establishment and monitoring point layout Relying on the Beijing Metro Shilihe Station as the background, a model of the new subway line 17 interchange channel was established to closely penetrate the existing subway line 10 Shilihe Station in a concealed tunnel, As shown in Fig. 11.The upper level of the existing Metro Line 10 tunnel is simplified as a horseshoe-shaped structure, with a width of 16 m and a height of 15 m.The new Metro Line 17 interchange has been simplified as a two-span, flat-roofed, straight-walled rectangular structure, with a width of 16 m and a height of 9 m. It features 1-meter-thick top and bottom plates, 1-meter-thick side walls, and a 1-meter-thick center wall, and is situated adjacent to the existing Metro station.The structural parameters of the station are shown in Table 3. Calculation model. Table 3Parameters of the supporting structure model.Serial numberNameElastic Modulus(MPa)Poisson’s ratioDensity(kg/m3)1Primary concrete C2528,0000. 2024002Main structure concreteC3030,0000. 202500 Parameters of the supporting structure model. According to the geological survey report, the soil layers are divided into 4 layers according to their nature and mechanical parameters, as shown in Table 4. Table 4After simplifying the soil parameters.Nameρ(g/cm3)c(KPa)φ(°)νE(MPa)Plain fill1. 878. 010. 00. 303. 0Clayey Silt2. 050. 028. 00. 3220. 0Silty fine sand2. 0130. 022. 00. 2711. 0Silty clay2. 2835. 024. 00. 3616. 0 After simplifying the soil parameters. In order to analyze the vibration response and deformation characteristics during the construction of the new interchange tunnel, under the influence of train vibrations, the model arranges monitoring sections along the excavation direction of the new interchange tunnel. In the closely-packed penetration section, one monitoring section is set up every 3.5 meters, labeled as ① through ④ and ①’ through ④’, respectively. In the non-closely-packed penetration section, one monitoring section is established every 5 meters, labeled as ⑤ through ⑩ and ⑤’ through ⑩’, respectively. Each of these sections is equipped with eight monitoring points at the same locations as those used in the field monitoring, specifically A1, A2, A3, A4, B1, B2, B3, B4. as shown in Fig. 12. Monitoring location layout. Model boundary conditions and dynamic parameters The dimensions of the model developed in this study should be large enough so that the external boundaries can represent an infinitely extended medium, i.e., both horizontal and vertical displacements vanish at both lateral boundaries.Therefore, it is established that the existing Line 10 tunnel is 40 m away from both boundaries, the new interchange station is 30 m away from both boundaries, and the bottom of the model is 30 m away from the bottom plate of the interchange station.The following assumptions were made to facilitate the numerical simulation calculations: Rocks and soils generally exhibit anisotropic behavior. The stratigraphic representation is simplified based on the geological survey report. The effect of groundwater hydrostatic pressure is not considered in the model. Initial in-situ stresses are considered only for soil self-gravitational stresses, while tectonic stresses in rocks and soils are neglected. Rocks and soils generally exhibit anisotropic behavior. The stratigraphic representation is simplified based on the geological survey report. The effect of groundwater hydrostatic pressure is not considered in the model. Initial in-situ stresses are considered only for soil self-gravitational stresses, while tectonic stresses in rocks and soils are neglected. It is assumed that the new underground test tunnels, existing subway station structures and soils comply with the deformation coordination principle. It is assumed that the new underground test tunnels, existing subway station structures and soils comply with the deformation coordination principle. The deformation of the track structure is assumed to be consistent with the deformation of the existing station and the train vibration model is applied to the station floor; in the analysis, an elastic deformation model is used for the lining and a Mohr’s Coulomb deformation model is used for the soil. The deformation of the track structure is assumed to be consistent with the deformation of the existing station and the train vibration model is applied to the station floor; in the analysis, an elastic deformation model is used for the lining and a Mohr’s Coulomb deformation model is used for the soil. In order to accurately describe the wave propagation in the model, the dimensions of the model cells are all smaller than 1/10 to 1/8 of the wavelength corresponding to the highest frequency of the input wave. In order to accurately describe the wave propagation in the model, the dimensions of the model cells are all smaller than 1/10 to 1/8 of the wavelength corresponding to the highest frequency of the input wave. In the initial static equilibrium calculation, the model temporarily turns off the dynamic calculation function. In this calculation, the original time-domain waveform is filtered and its first-order integral is reduced to a dynamic load in the form of a sinusoidal waveform applied to the bottom of the existing station. The applied dynamic load consists of three parts: the first part is applied along the model in the Y-direction (along the existing station) with a peak acceleration of 2 m/s2; the second part is applied in the X-direction with a peak acceleration of 1 m/s2; and the third part is applied in the Z-direction with a peak acceleration of 0.8 m/s2. The dynamic wave lasts for 6 m/s2. In FLAC3D, the dynamic load is usually entered into the model as a “readout” in the form of a FISH function which creates the acceleration time history and stores it in the table at 0. 002s time intervals. The acceleration equation for the train vibration wave and its time history curve are shown in the following Eq. (1) The new interchange is a field-shaped frame structure, which is mainly composed of I-beams.There are three types of I-beams in common use: I16,I18 and I20b.Five spacing options were selected for the I-beam arrangement: 0.8 m, 1.6 m, 2.4 m, 3.2 m and 4.0 m.The placement of different types and spacing of I-beams affects the structural stiffness, resulting in different vibration response characteristics, and the results of the study have some engineering value. Model validation This model verifies that the initial support structure of the interchange channel is made of I18 I-beams with a spacing of 3.2 m. Simulating the real running speed of trains entering and leaving the station, the speed is about 40 km/h. Substituting this condition into the numerical simulation model for calculation, the result is shown in Fig. 13. Horizontal and vertical acceleration plotted against time. (a)~(h) Indicates the monitoring situation of monitoring points A1 ~ A4 and B1 ~ B4. Comparison of the simulation results in Fig. 13 with the measured data in Fig. 9 shows that the acceleration time course curve obtained from the simulation follows the same trend as the measured data, with a difference of 6% in the peak acceleration value.The reason for this error is that the presence of plastic flow in the Mohr-Coulomb model consumes a lot of energy.Compared to the elastic model, the additional damping and the choice of its parameters have a greater influence on the calculation results.In this model, the soil is regarded as a cohesive material, where the shear modulus exhibits hysteresis, decaying with shear strain. The related hysteresis damping parameters default to the fitting process of the model. As a result, some high-frequency noise remains unremoved. Take the maximum vibration level (VLmax) as the evaluation index, analyze the change rule of the vibration acceleration of the frame structure at each position caused by the train operation, as shown in Fig. 14. As the distance of the frame structure from the vibration source increases, the vertical and horizontal accelerations tend to decrease, which is influenced by the radiation damping of the vibration wave and the material damping of the frame structure.In addition, the horizontal vibration wave results for both walls in the same frame are inconsistent, with significant “vibration amplification” in the outer wall (A2). Vibration index. In order to further investigate the cause of the “vibration amplification” phenomenon in this test, we analyze the vibration response of each measurement point by converting it to the frequency domain through Fourier transform, as shown in Fig. 15. From the results, it can be seen that the main frequency components of horizontal acceleration at the frame monitoring points are concentrated in the intervals of 13–20 Hz, 30–36 Hz, 53–60 Hz and 79–87 Hz, and the main frequency components of vertical acceleration are concentrated in the intervals of 7–14 Hz, 22–27 Hz, 43–51 Hz and 65–69 Hz, and these frequency bands are in line with the characteristic frequencies measured in the field.From Fig. 15a to h, it can be seen that as the train vibration wave propagates inside the frame structure, the horizontal acceleration vibration bands in the intervals of 13 ~ 20 Hz and 53 ~ 60 Hz show a rapid weakening trend.The vibration frequency in the interval of 30 ~ 36 Hz and 79 ~ 87 Hz shows a tendency of increasing and then decreasing, and reaches the maximum value at the measurement point of A2, which again proves the existence of obvious “vibration amplification” phenomenon in the outer wall (A2).Compared with the vibration frequency of horizontal acceleration, the vibration of each frequency band of vertical acceleration exhibits a rapid attenuation trend. Specifically, within the frequency intervals of 7 ~ 14 Hz and 65~69 Hz, the vibration wave of the train has essentially completed its basic attenuation by the time it reaches the side wall (A2, A4). Horizontal and vertical acceleration versus frequency.(a)~(h) Indicates the monitoring situation of monitoring points A1 ~ A4 and B1 ~ B4. Results Impact of different types of I-beams Vibration acceleration In order to study the vibration response characteristics of various types of I-beams in the structure of the new interchange, the study selected three types of I-beams for modeling: I16, I18, and I20b. The spacing between the I-beams was set at 3.2 m. For the simulation calculations, the study considered four train speed conditions: station deceleration, acceleration, and full-speed passing at 20 km/h, 40 km/h, 60 km/h, and 80 km/h.The point peak acceleration (PPA) values at each monitoring point within monitoring section ① (center of the dense patch), monitoring section ③ (center of the non-dense patch), and monitoring section ⑤ (non-dense patch area) were extracted and are presented in Fig. 16. Plot of PPA changes at each monitoring point when using different I-beams.(a)~(d) represents the train speeds of 20, 40, 60, and 80 km/h, respectively. From the analysis of Fig. 16a to d, it can be observed that the vibration acceleration of the frame structure within the interchange channel increases as the train running speed increases; Within the same cross-section, the vibration acceleration of the frame structure of the interchange channel decreases sequentially from top to bottom. However, there is an inflection point at point A2, leading to the phenomenon known as the ′amplification effect′. This effect is caused by the fact that the vibration attenuation of the soil body is weaker than that of the structural metal material. Consequently, the vibration wave is superimposed onto the structure twice after the initial vibration of the soil body.The vibration acceleration value of the lower guideway of the interchange passage is small, and the impact of train vibration on the lower guideway is small, which is consistent with the measured situation ;the same conditions and the same cross-section, the selection of I16 I-beam than I20b I-beam point peak acceleration (PPA), larger I-beam cross-section leads to greater structural rigidity, vibration resistance effect is more obvious. By simulating the effects of different train speeds and I-beam types on the vibration of the interchange frame structure, the vibration signals of monitoring points A1, A2 and A4 in section ① in model Fig. 12 were extracted and decomposed, and the frequency-amplitude relationship Fig. 17 was obtained. As illustrated in Fig. 17a, during low-speed operation at a vehicle speed of 20 km/h, two peaks are observed at monitoring point A1 in the closely packed area of the frame structure utilizing I16 I-beams. These peaks are distributed within the frequency ranges of 32–38 Hz and 72–88 Hz, respectively. In contrast, under other operating conditions, only a single peak is present.From Fig. 17a to d, it can be observed that as the train speed increases, the number of frequency intervals where significant vibrations occur also increases. Specifically, the amplitude reaches peak values within four distinct frequency ranges: 12–18 Hz, 30–40 Hz, 52–56 Hz, and 72–88 Hz.;When the train speed is 20 km/h, the vibration acceleration at point A2 using I16 I-beam is 0.0003 m/s² greater than at point A4, and using I18 I-beam, the vibration acceleration at point A2 is 0.0002 m/s² greater than at point A4. At a train speed of 80 km/h, the frequency amplitude at point A2 using I16 I-beam is 0.0012 m/s² greater than at point A4, and using I18 I-beam, the frequency amplitude at point A2 is 0.0004 m/s² greater than at point A4. It can be seen that as the train speed increases, the vibration amplification effect becomes more pronounced, and as the stiffness of the frame structure increases, the vibration amplification effect diminishes. Horizontal acceleration versus frequency when using different I-beams. (a)~(d) represents the train speeds of 20, 40, 60, and 80 km/h, respectively. Displacement response As shown in Fig. 18, the final deformation profile of the new interchange tunnel exhibits a V-shape, with the maximum settlement occurring directly under the existing line. The settlement gradually decreases as one moves away from the existing line and eventually stabilizes at a distance of 25 m from the existing line, indicating that it is no longer influenced by loads (including both self-weight and vibration loads) beyond this point.The settlement value of the new interchange channel directly under the train is the largest, comparing the initial support with different I-beam I20b, I18, I16 as support, the maximum deformation in the top plate of the upper guide hole is 7.8 mm, 9.2 mm, 13.1 mm, respectively, and the selection of I20b I-beam is the smallest settlement in the close part of the I20b I-beam.Therefore, in the project, I-beam I20b can be used to support, and at the same time, appropriate reinforcement of the close-packed parts to prevent excessive settlement. Displacement response of different I-beam types. I-beam different spacing influence Vibration acceleration In order to study the vibration response characteristics of the new interchange structure with varying I-beam spacings, the I18 type I-beam is selected for modeling. The spacings considered for the I-beams are 0.8 m, 1.6 m, 2.4 m, 3.2 m, and 4.0 m. The train speeds are set to simulate deceleration, acceleration, and full-speed passage through the station, with four working conditions: 20 km/h, 40 km/h, 60 km/h, and 80 km/h. Simulation calculations are then performed. By extracting the peak point acceleration (PPA) data from each monitoring point in monitoring sections ① (center of dense patch), ③ (center of non-dense patch), and ⑤ (non-dense patch area), as shown in Fig. 19 . Plot of PPA changes at each monitoring point using different I-beam spacings. (a)~(d) represents the train speeds of 20, 40, 60, and 80 km/h, respectively. As shown in Fig. 19, the effect of different I-beam spacing on the vibration acceleration of the new structure is as follows: (1) When selecting the monitoring section ① at the full-speed station approach, it is observed that, with an I-beam spacing of 4.0 m, the maximum acceleration of 0.0172 m/s² occurs at the top of A1 within the initial support structure of the interchange tunnel. Conversely, when the I-beam spacing is reduced to 0.8 m, the maximum acceleration decreases to 0.0046 m/s² at the same location. This indicates that, as the spacing of the I-beams increases, the vibration response of the structure becomes more pronounced. Additionally, monitoring sections ③ and ⑤ exhibit similar trends. (2) From the monitoring section ①, it is evident that the larger the I-beam spacing, the more pronounced the vibration amplification effect becomes. When the I-beam spacing is 4.0 m, the peak particle acceleration (PPA) of A2 and A4 is 0.0116 m/s² and 0.0072 m/s² respectively, resulting in an amplification value of 0.0044 m/s² and an amplification rate of 61.1%. Conversely, when the I-beam spacing is reduced to 0.8 m, the PPA of A2 and A4 decreases to 0.0028 m/s² and 0.0022 m/s² respectively, with an amplification value of 0.0006 m/s² and an amplification rate of 27.3%. Meanwhile, monitoring sections ③ and ⑤ exhibit similar trends. (3) The amplification effect diminishes as the distance from the vibration source increases. For instance, in monitoring section ⑤ (a non-dense area), when the I-beam spacing is set to 4.0 m, the peak particle acceleration (PPA) amplification values for A2 and A4 are 0.0036 m/s² and 0.0042 m/s² respectively. The overall amplification value for this section, considering both A2 and A4, might be derived as 0.0006 m/s², yielding an amplification rate of 16.7%. Consequently, the vibration amplification effect in this area is not pronounced. The vibration signals from monitoring points A1, A2, and A4 in section ① of the model depicted in Fig. 12 have been extracted and decomposed. The resulting frequency-amplitude relationship is illustrated in Fig. 20a. When the vehicle operates at a low speed of 20 km/h, two peaks are observed in the vibration signal at point A1, situated at the closely packed area of the frame structure where I-beams are spaced 4 m apart. These peaks fall within the frequency ranges of 32–38 Hz and 76–84 Hz. Under other working conditions, only one peak is noted at point A1.From Fig. 20a, b, c, and d, a clear pattern emerges regarding the frequency amplitudes at monitoring points A2 and A4. Specifically, when the train operates at a speed of 20 km/h and the I-beam spacing is 0.8 m, the frequency amplitude at A2 is 0.00008 m/s² higher than at A4. As the I-beam spacing increases to 4 m (while maintaining the same train speed), the difference in frequency amplitude between A2 and A4 enlarges to 0.0006 m/s². Similarly, when the train speed rises to 60 km/h with an I-beam spacing of 0.8 m, the frequency amplitude at A2 exceeds that at A4 by 0.00018 m/s². At an I-beam spacing of 4 m and a train speed of 60 km/h, the difference further widens to 0.0008 m/s². Overall, these observations indicate that both an increase in I-beam spacing and an increase in train speed contribute to a more pronounced vibration magnification effect. lot of acceleration versus frequency at each monitoring point using different I-beam spacings. (a)~(d) represents the train speeds of 20, 40, 60, and 80 km/h, respectively. Displacement response As shown in Fig. 21, when the I-beam spacing is 0.8 m, the maximum settlement generated at the location A1 of the top plate of the upper guide hole is 7.2 mm; when the I-beam spacing is 1.6 m, 2.4 m and 3.2 m, the maximum settlement generated is 11.6 mm, 11.7 mm and 11.8 mm respectively, and the amount of settlement is basically the same.When the I-beam spacing is set at 4.0 m, the maximum settlement at location A1 is 13.8 mm. From this, it can be observed that the denser the initial support I-beam arrangement in the interchange tunnel, the greater the overall structural rigidity, the better the vibration-resistant performance, and the smaller the structural settlement will be. Displacement response of different I-beam spacing. Effect of train location Vibration acceleration In order to further study the vibration response characteristics of the frame structure below the close center, the vibration response of the frame structure of the interchange tunnel will become more intricate when considering the subway train passing through the proximal end continuously from the far end until the close center.Therefore, in this study, every 4 m along the direction of train travel is designated as one position. There are a total of 10 positions considered, specifically: when the train has not yet arrived at the far end of the frame structure (position 5); when the train passes through the boundary of the frame structure at the near end (position 3); and when the train is centered directly below the top of the frame structure (position 1). The corresponding monitoring points on the frame structure are labeled as 1-A1, 1-A2, 1-A4, …, up to 5-A1, 5-A2, 5-A4, and so forth, as illustrated in Fig. 22. Schematic layout of monitoring sections. (a) The train travels to a position far from the close position. (b) The train travels to a position near from the close position. (c) The train is moving to the center of the close position. As shown in Fig. 23, when the train reaches the position directly above the frame structure (referred to as the “top train” position), the overall amplitude at the A1 position within the frame structure is the largest. The amplitude increases and then decreases within the frequency ranges of 12–20 Hz, 30–36 Hz, 53–56 Hz, and 60–81 Hz, with peak amplitudes occurring at 15 Hz, 35 Hz, 55 Hz, and 80 Hz. Among these, the peak amplitude at 80 Hz is the largest. When the train arrives at different positions along its travel path, the amplitudes at the A2 and A4 positions both exhibit obvious inconsistencies. Specifically, the peak amplitude at the outer wall of A2 is larger than the peak amplitude at the inner wall of A4. For instance, when the train is at the far end of its travel path (relative to the frame structure), the difference between the peak amplitudes of A2 and A4 is 0.00021 m/s²; whereas when the train is at the “top train” position, the difference increases to 0.00251 m/s².It can be seen that the closer the train is to the frame structure, the more pronounced the vibration response is. Horizontal acceleration versus frequency. Displacement response Figure 24 shows the deformation values of each monitoring point A1-A4, B1-B4 when the train passes through different positions and the corresponding frame structure displacement cloud of some train positions.From Fig. 24a, it can be seen that when the train runs to the top of the frame structure, the vibration leads to the largest displacement response of the frame structure as a whole. Among these, the displacement response of the top plate (A1) inside the frame structure is the largest, reaching 9.6 mm. From Fig. 24b, it can be observed that as the distance to the frame structure increases, the displacement response of the frame structure tends to zero at position 6. This indicates that the farther the train is from the frame structure, the smaller the corresponding displacement response. Comparing the vertical displacements of the frame structure at different train positions, there is an obvious inconsistency in the displacements at positions A2 and A4 from the proximal train to the top train. Specifically, at the top train position, A2 and A4 in the frame structure exhibit displacements of 7.9 mm and 6.8 mm, respectively. Combined with the displacement cloud diagrams, this suggests that the side of the frame structure close to the soil body is more affected by the vibration. Change in displacement of the frame structure as the train passes through different locations.(a)Status of monitoring sites(b)Surrounding area of influence. Discussion Equivalent stiffness theory From the previous study, it is found that the frame structure composed of I-beams and surrounding soil will produce vibration amplification effect when subjected to train vibration.In order to investigate the influence of different I-beam parameters on the strength of the vibration amplification effect, this study investigates by altering the I-beam parameters, such as the type of I-beam, the spacing of I-beams, as well as the running speed of trains on the existing line, and finds a positive correlation between the spacing of I-beams in the frame structure (denoted as B), the running speed of the trains on the existing line (denoted asV), and the range of influence f(L).And the modulus of elasticity function I of the I-beam shows a negative correlation with the range of influence f(L).By introducing the correction factor α, the functional relationship between the influence range L of the soil layer adjacent to the frame structure and the operating speed V of the train on the existing line and the parameters B and I of the I-beam is given in the following equation: Where f(L) is the affected range of the adjacent soil layer of the frame structure, α is the correction factor, B is the spacing of the frame structure I-beam, V is the running speed of the existing line train, I is the modulus of elasticity of different kinds of I-beam. Change the I-beam spacing B and I-beam type I is equivalent to the change in I-beam stiffness, the equivalent stiffness Q is: Where Q is the equivalent stiffness of I-beam, n is the correction factor. By fitting the previous numerical simulation results, when the I-beam spacing is 2.4 m, the functional relationship between the type of I-beam and the affected range of the soil layer adjacent to the frame structure is obtained as: When the I-beam type is I18, the spacing of the I-beam as a function of the extent of influence on the soil adjacent to the frame structure is: When the I-beam type is I18 and the I-beam spacing is 2. 4 m, the relationship between the operating speed of the existing line trains and the affected range of the soil adjacent to the frame structure is: : By iteration we get: Similarly, under the conditions of different I-beam parameters and different established line train speeds, the vibration amplification effect will produce certain settlement changes to the soil layer adjacent to the frame structure. By introducing the correction coefficient β, the maximum settlement f(S) of the soil layer adjacent to the frame structure as a function of the train speed V and the I-beam parameters is as follows: Where: f(S) is the maximum settlement of the soil adjacent to the frame structure, β is the correction factor. When the I-beam type is I18 and the I-beam spacing is 2.4 m, the relationship between the operating speed of the existing line train and the maximum settlement of the soil adjacent to the frame structure is: By iteration we get: Vibration amplification effect In the previous study, it was found that during the train vibration process, the vibration frequencies on the two sides of the frame structure were significantly inconsistent, with the vibration frequency on the side close to the soil being greater than that on the inner side of the frame. Figures 25 and 26 illustrate the deformation of the soil and structure under different I-beam types and spacings when trains passed through the top of the frame structure of the interchange at speeds of 20 km/h, 40 km/h, 60 km/h, and 80 km/h, respectively. Observations from Figs. 25 and 26 reveal: Settlement of different I-beam types. (a)~(d) represents the train speeds of 20, 40, 60, and 80 km/h, respectively. Settlement of different I-beam spacing. (a)~(d) represents the train speeds of 20, 40, 60, and 80 km/h, respectively. (1)When the train passes through the top of the interchange, the deformation at the frame structure of the interchange is the largest under different I-beam types and spacing, and the deformation of the surrounding soil is relatively small, with an overall V-shaped trend; (2)When the train passes through the top of the interchange, an inflection point emerges in the settlement profile at the junction of the frame structure and the adjacent soil layer. As the train speed increases, the I-beam cross-section decreases, and the I-beam spacing widens, the inflection point becomes more prominent and the mutation value increases; (3) From Eqs. (2), (3), (11), (12), it can be seen that the speed of the existing line train is V ≤ 20 km/h, change the parameters of the frame structure I-beam, the equivalent stiffness of the I-beam Q ≥ 1, the settlement range of L ≤ 1. 8 mm, the frame structure adjacent to the soil body has no vibration amplification effect, at this time, do not need to consider the settlement control. When the speed is 20 ≤ V ≤ 40 km / h, I-beam equivalent stiffness Q ≥ 0.8, or the speed of the existing line train is 40 ≤ V ≤ 60 km / h, I-beam equivalent stiffness Q ≥ 0.5, settlement range of 1.2 ≤ L ≤ 1. 8 m, the maximum amount of settlement is 3.6 mm, there is weak vibration amplification, this time, it should be mainly for the guideway roof to take measures to control settlement. When the running speed of the existing line train is 60 ≤ V ≤ 80 km/h, the equivalent stiffness of I-beam Q ≥ 0.2, change the frame structure I-beam parameters, settlement range 1. 8 ≤ L ≤ 3. 6 m, maximum settlement 9. 6 ≤ S ≤ 17. 8 mm, there is a strong vibration amplification effect, at this time, should be on the roof plate of the guide hole and the inner and outer sidewalls to take measures to control settlement. Vibration propagation From the previous study, the deformation of the frame structure is mainly concentrated in the top plate of the upper level, and the deformation of the frame frame structure of the lower level is not large; because it is a close-fitting structure, the deformation of the top plate of the upper level of the frame structure directly affects the running track of the existing line subway track on the upper level. The parameters of the frame structure remain constant. The faster the speed of the train running in the proximity to the tunnel wall, the greater the vertical deformation at the center of the top plate, taking on a ‘V’ shape due to the deformation. The stronger the contact between the wheels of the subway train and the train track, the more intense the vibration response transmitted through the frame structure, which is directly influenced by the vibration source . When the train speed is kept constant, changing the I-beam parameters of the frame structure results in different deformations. Specifically, when the I-beam type is I20b and the I-beam spacing is 0.8 m, the deformation of the frame structure is smaller. Conversely, if the I-beam spacing is larger or the I-beam strength is too small, it will adversely affect the overall strength of the frame structure.When the upper train operation generates vibration loads that are applied to the frame structure, the deformation will be larger. Conversely, when the spacing and strength of the I-beams are appropriately set, the deformation will be smaller. Furthermore, when the overall strength of the frame structure is sufficient, different spacing and types of I-beams will influence the vibration resistance properties of the frame structure, ultimately affecting its deformation. This deformation, in turn, is influenced by the propagation path of the vibrations. From Fig. 27, it can be observed that when there is no soil on both sides of the frame structure, the vibrating load is transferred solely through the frame structure to the lower base. However, when soil is present on both sides, the vibrating load can be transferred not only to the lower base but also laterally to the soil on both sides. As a result, the vibration load on the soil on both sides increases. A vibration amplification area emerges at the boundary between the frame structure and the soil layer, and this area is concentrated near the upper layer of the frame structure. Vibration Transmission Schematic. (a) Vibration transmission path when there is no lateral soil layer. (b) Vibration transmission path in the presence of lateral soil layers. From Fig. 28a, b, c, d it can be observed that the train speed increases, the vibration amplification area within the dashed portion of the rectangle becomes increasingly prominent. Based on previous studies, only horizontal vibrations exhibit a vibration amplification phenomenon, while vertical vibration waves undergo refraction and amplification as they propagate to the interface between the frame structure and the soil layer. Based on numerical simulation results, the vibration amplification areas corresponding to different train speeds, train vibration acceleration, I-beam spacing, and I-beam models are plotted in Fig. 28e, f. It can be seen from the figure that the speed of the train passing through the frame structure decreases, the vibration generated by the train causes less vibration in the structure and surrounding soil environment, and the vibration amplification area becomes smaller. For I16, I18 and I20b, the vibration amplification effect range generated by I-beams using the I20b model is the smallest. Among them, I20b has the largest modulus of elasticity and I16 the smallest; when the modulus of elasticity is large, the vibration load attenuates more when it is transmitted in the frame structure, and the vibration load is smaller when it is transmitted to the soil on both sides. For different I-beam spacing, 0.8 spacing is the smallest settlement, and the smaller the range of vibration amplification effect. Considering the frame structure as a whole, changing the spacing of I-beam is to change the overall stiffness of the frame structure. The greater the stiffness, the more the vibration load is attenuated when it is transmitted in it, and the load transmitted to the soil on both sides is smaller, which leads to a smaller range of vibration amplification effect. Vibration amplification influence diagram.(a) 20 km/h vibration amplification cloud map. (b) 40 km/h vibration amplification cloud map. (c) 60 km/h vibration amplification cloud map. (d) 80 km/h vibration amplification cloud map. (d) Velocity and vibration acceleration. (e) I-beam spacing and type. Conclusion Based on the new Beijing Metro Line 17 project, which closely intersects with the existing line, a specific study was conducted through on-site monitoring and numerical simulation. The main content of the study is to analyze the dynamic response and support effectiveness of the closely adjacent interchange station under train vibration loads. The main conclusions of the study are as follows: (1) The numerical calculation results are basically consistent with the actual measured values.When the train load passes through, the time-domain waveform has obvious peaks. The acceleration response inside the close-fitting frame structure decreases with increasing distance from the closely packed area, and the vibration primarily affects the upper frame structure, with minimal impact on the lower part; In frame structures, the vertical acceleration of vibration decreases with distance from the source of vibration; However, the vibration response in terms of horizontal vibration acceleration at the junction between the frame structure and the soil body will suddenly increase, resulting in a vibration amplification effect. Along the excavation direction horizontally, the frame structure that is not closely adhered to the tunnel wall exhibits the same trend as the closely adhered frame structure. However, the amplitude of vibration in the non-closely adhered frame structure gradually decreases as the horizontal distance increases, indicating an obvious attenuation phenomenon in energy transfer. At the same time, the vibration amplification effect appears to be significantly weakened. (2) There are three reasons for the settlement of the close-fitting station: the compression settlement of the soil is the largest, followed by the deflection deformation of the existing structure, and the elastic deflection deformation of the second lining of the new structure is the smallest. The main performance of flat-topped and straight-walled stations is the overall vertical displacement of the structure under the action of train load. The settlement of the sticking part is smaller than that of the non-sticking part, but the displacement difference at the junction of the sticking part is obvious. (3) For different I-beams commonly used in the project, the displacement curves of different I-beams arranged at the close contact position showed the same trend of the displacement curves: first increasing, then decreasing, and then tending to equilibrium. Then the I-beams were arranged at different spacings, and it was found that the 20b I-beams with a spacing of 0.8 m were the most suitable. (4) Changing the type and spacing of the materials that closely fit the frame structure only controls the internal vibration response of the frame structure. The vibration response at the boundary between the frame structure and the soil is not controlled, which affects the safety of the structure. Acknowledgements This study was funded by the National Natural Science Foundation of China (Grant Nos.52378396). Author contributions Yunsi Liu (First Author; Corresponding Author): Conceptualization, Software, Investigation, Methodology, Writing–original draft, Writing–review & editing; Yueyuan Huang (Second author): Software, Investigation, Methodology, Writin–original draft, Writing–review & editing. Yu Huang (Author 3):Writing–review & editing. Yonglin An (Author 4):Visualization, Investigation; Yaxiong Peng (Author 5):Resources, Supervision. Data availability The authors confirm that the data supporting the findings of this study are available within the article. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Yunsi Liu, Email: liuyunsi@sohu.com. Yu Huang, Email: huangyu2022@sohu.com. References Associated Data Data Availability Statement The authors confirm that the data supporting the findings of this study are available within the article.