Method for simulating and analyzing near-field noise modal energy distribution of pantograph of high-speed train

Abstract

The invention discloses a method for simulating and analyzing near-field noise modal energy distribution of a pantograph of a high-speed train. The method comprises the following steps: S1) establishing a three-dimensional finite element model of the high-speed train with the pantograph and carrying out grid division; s2) performing fluid simulation analysis on the grid model established in the step S1); s3) collecting sound pressure data at grid nodes on a plurality of sections which are located at different space heights of the pantograph area and are parallel to the ground; and S4) processing and analyzing the acquired sound pressure data by adopting an SPOD method to obtain a noise modal energy distribution diagram between SPOD modal energy and frequency. Through the method, a noise modal energy distribution diagram at each section of the pantograph area of the high-speed train can be obtained through simulation analysis, so that near-field sound pressure distribution and spatial propagation characteristics of the pantograph can be obtained, and a theoretical basis can be provided for near-field pneumatic noise reduction of the pantograph of the high-speed train; the method has important significance and application value for solving the aerodynamic noise of the high-speed train.

Description

Technical Field

[0001] The invention relates to a method for simulating and analyzing the near-field noise modal energy distribution of a high-speed train pantograph, and belongs to the technical field of aerodynamic noise analysis. Background Art

[0002] In recent years, with the continuous development of high-speed rail technology, the operating speed of high-speed trains has been continuously increasing. By the end of 2023, the operating mileage of my country's high-speed rail has reached 45,000 kilometers, and the maximum operating speed has reached 350km. However, existing studies have shown that when the speed of a high-speed train exceeds 300km / h, the aerodynamic noise on the train will exceed the wheel-rail noise and become the main noise source. Among them, the pantograph area is one of the main aerodynamic noise sources on the high-speed train, and the pantograph near-field noise is part of the pantograph aerodynamic noise. It is caused by the sound radiation part of the aerodynamic excitation, specifically the noise generated by the near-field sound pressure change in the pantograph area. With the increase in the speed of the train, the near-field aerodynamic noise caused by the pantograph area also increases significantly, which will significantly affect the comfort of the train crew and passengers. However, in order to reduce the near-field aerodynamic noise of the pantograph of a high-speed train, it is necessary to first study the generation mechanism and distribution characteristics of the near-field aerodynamic noise of the pantograph of a high-speed train.

[0003] The modal energy distribution of aerodynamic noise refers to the energy distribution of aerodynamic noise at different frequencies. In the study of aerodynamic noise, modal decomposition is an important tool that helps us better understand the generation and propagation mechanism of noise. By analyzing the modal energy distribution of pantograph near-field noise, we can understand the spatial transfer characteristics of near-field sound pressure at the pantograph, which can better help design noise control strategies and provide an important theoretical reference for reducing pantograph near-field aerodynamic noise. However, there has been no report on the research method of modal energy distribution of near-field noise of high-speed train pantographs. Summary of the invention

[0004] In view of the above problems and requirements in the prior art, an object of the present invention is to provide a method for simulating and analyzing the near-field noise modal energy distribution of a high-speed train pantograph.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] A method for simulating and analyzing the near-field noise modal energy distribution of a high-speed train pantograph comprises the following steps:

[0007] S1) Establish a 3D finite element model of a high-speed train with a pantograph and perform meshing;

[0008] S2) performing fluid simulation analysis on the grid model established in step S1);

[0009] S3) collecting sound pressure data at each grid node on several cross sections located at different spatial heights in the pantograph area and parallel to the ground;

[0010] S4) The collected sound pressure data are processed and analyzed using the spectral proper orthogonal decomposition (SPOD) method, that is, a noise modal energy distribution diagram between SPOD modal energy and frequency is obtained.

[0011] Furthermore, the step S1) includes the following specific operations:

[0012] S11) Create a three-car train model with the first and last cars and the pantograph car;

[0013] S12) Create a computational domain, including the air inlet and outlet, the bottom ground, and the symmetric boundaries of the remaining three surfaces;

[0014] S13) removing the train model in the computational domain created in step S12) and leaving only the part where the air exists, and then performing meshing.

[0015] In one implementation scheme, step S12) is to create a calculation domain in accordance with the national railway standard TB / T 3503.4-2018 "Specifications for Numerical Simulation of Train Aerodynamic Performance".

[0016] In one implementation scheme, in step S13), the trimmed body provided by STAR-CCM+ is used for meshing.

[0017] Furthermore, the step S2) includes the following specific operations:

[0018] S21) adopts k- The turbulence model is used to calculate the steady flow field;

[0019] S22) using the steady flow field calculated in step S21) as an initial condition, and performing unsteady flow field calculation using large eddy simulation (LES);

[0020] S23) After the flow field reaches dynamic equilibrium, the acoustic perturbation equation (APE) is used to calculate the acoustic field.

[0021] In one implementation, the k- The governing equations of the turbulence model are:

[0022] ;

[0023] Where: is the fluid density; k is the turbulent kinetic energy; t is the time; x j is the y-component of the vortex scale; u j is the y-component of the gas flow velocity; is the x-component of the turbulent viscosity; is the turbulent viscosity; is the eddy viscosity model of Reynolds stress; S ij is the average velocity strain rate tensor; β* is the model constant, which is taken as 0.09; is the internal dissipation rate of turbulent kinetic energy; For the generated items; , ,β,F1, are model constants.

[0024] In a preferred solution, the number of calculation steps in step S21) is set to 1000 steps.

[0025] In one implementation, the control equation of the large eddy simulation (LES) in step S22) is:

[0026] ;

[0027] Where: is the x-component of the gas flow velocity after filtering, is the y-component of the filtered gas flow velocity, and i≠j=0,1,2; t is time; x j is the y-component of the vortex scale; is the fluid density; is the pressure after filtering; x i is the x-component of the vortex scale; is the viscosity coefficient; is the sub-grid stress.

[0028] In a preferred solution, the number of calculation steps in step S22) is 1000, the time step is 0.0005s, and each step is iterated 10 times.

[0029] In one implementation, the acoustic perturbation equation (APE) in step S23) is:

[0030] ;

[0031] Where: c is the speed of sound; is the sound pressure; t is the time; is the time average speed; is the fluid dynamic pressure; is the damping coefficient; is the time step; , is the local Courant number, which reflects the movement distance of the fluid in the grid, where is the grid spacing, Delta is the increment. , x n and x n-1are two adjacent nodes in the grid; in addition, is the nabla operator, which is a vector differential operator. The specific calculation method is as follows:

[0032]

[0033] Among them: I, J, K represent the direction components along the positive directions of the three coordinate axes x, y, and z respectively.

[0034] In a preferred solution, the number of calculation steps in step S23) is 500 steps, the time step is 0.0001s, and each step is iterated 10 times.

[0035] In one implementation scheme, the boundary conditions for fluid simulation analysis in step S2) are set as follows: the inlet is a velocity inlet, which is set to 111 m / s (equivalent to simulating a train running at 400 km / h); the outlet is a pressure outlet, which is set to 0 Pa; the ground is a no-slip wall, and the remaining surfaces are symmetrical boundaries, simulating the train running in an open-air environment.

[0036] In a preferred embodiment, step S3) collects 1500 steps in total, and the time step is 0.0001 s.

[0037] In one implementation scheme, the specific steps of processing and analyzing in step S4) are as follows:

[0038] S41) importing the sound pressure data collected in step S3) into MATLAB software in the form of compressed snapshots;

[0039] S42) Construct the snapshot matrix as Q=[q1,q2,…,q m ], where q m is an n-row, 1-column matrix containing the sound pressure information of each grid node on the cross section at time m, where n is the number of grid nodes on the cross section;

[0040] S43) The snapshot matrix constructed in step S42) is divided into a plurality of overlapping block matrices by using the Welch method, and a block matrix q with N0 adjacent overlapping snapshots is obtained. Nf , where Nf is the number of snapshots included in each block matrix;

[0041] S44) Perform discrete Fourier transform and windowing on each block matrix, and then extract the specified frequency f g The snapshots under are rearranged to obtain a new block matrix, and then the Fourier components of each block are arranged according to frequency to form a new snapshot matrix;

[0042] S45) solve the eigenvalue problem of each frequency correlation matrix, that is, obtain the SPOD mode and the corresponding modal energy at each eigenfrequency, and then sort the modes according to the size of the modal energy to obtain the relationship diagram between the modal energy of each order and the frequency, so as to obtain the noise modal energy distribution diagram at each section.

[0043] Compared with the prior art, the beneficial technical effects of the present invention are:

[0044] The method of the present invention can be used to simulate and analyze the noise modal energy distribution diagram at each cross section of the pantograph area of ​​a high-speed train. From the obtained noise modal energy distribution diagram, the change of the peak frequency of each order modal energy and the frequency range of the main energy distribution can be known, and the change trend of the frequency range with the spatial position height can be known, so as to obtain the near-field sound pressure distribution and spatial propagation characteristics of the pantograph, thereby providing a theoretical basis for the near-field aerodynamic noise reduction of the pantograph of a high-speed train, which is of great significance and application value for solving the aerodynamic noise of high-speed trains. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a train model diagram created for simulation analysis in an embodiment;

[0046] Figure 2 is a schematic diagram of a calculation domain used for simulation analysis in an embodiment;

[0047] Figure 3 is a cross-sectional mesh division diagram of the computational domain in the embodiment;

[0048] Figure 4 is a cross-sectional position structure diagram collected by simulation analysis of the embodiment;

[0049] Figure 5 is a noise modal energy distribution diagram of the bow head cross section obtained by simulation analysis in the embodiment;

[0050] Figure 6 is a noise modal energy distribution diagram of the cross section of the arm connection obtained by simulation analysis of the embodiment;

[0051] Figure 7 It is a noise modal energy distribution diagram of the bottom surface of the pantograph cavity obtained by simulation analysis of the embodiment. DETAILED DESCRIPTION

[0052] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. In addition, it should be noted that the terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. Unless otherwise defined, the technical terms or scientific terms used in the present invention should be the usual meanings understood by people with general skills in the field. In addition, in the embodiments, the direction parallel to the center line of the track is the X-axis direction, the direction horizontally perpendicular to the center line of the track is the Z-axis direction, and the direction vertically perpendicular to the center line of the track is the Y-axis direction.

[0053] Example

[0054] A method for simulating and analyzing the near-field noise modal energy distribution of a high-speed train pantograph comprises the following steps:

[0055] S1) Establish a three-dimensional finite element model of a high-speed train with a pantograph and perform meshing. The specific operations are as follows:

[0056] S11) Create a three-car train set model with the first and last cars and the pantograph car, see Figure 1 As shown;

[0057] S12) Create a computational domain in accordance with the national railway standard TB / T 3503.4-2018 "Numerical Simulation Specification for Train Aerodynamic Performance", including the air inlet and outlet, the bottom ground and the symmetrical boundaries of the other three surfaces, see Figure 2 As shown, H in the figure represents the height value of the train;

[0058] S13) Remove the train model in the computational domain created in step S12) and leave only the part where the air exists. Then use the trimmed volume provided by STAR-CCM+ to perform meshing, and obtain Figure 3 The cross-sectional meshing diagram of the computational domain is shown;

[0059] S2) Perform fluid simulation analysis on the grid model established in step S1), and the specific operations are as follows:

[0060] S21) adopts k- The turbulence model is used to calculate the steady flow field, and the number of calculation steps is set to 1000. The governing equations of the turbulence model are:

[0061] ;

[0062] Where: is the fluid density; k is the turbulent kinetic energy; t is the time; x j is the y-component of the vortex scale; u j is the y-component of the gas flow velocity; is the x-component of the turbulent viscosity; is the turbulent viscosity; is the eddy viscosity model of Reynolds stress; S ij is the average velocity strain rate tensor; β* is the model constant, which is taken as 0.09; is the internal dissipation rate of turbulent kinetic energy; For the generated items; , ,β,F1, are model constants;

[0063] S22) Taking the steady flow field calculated in step S21) as the initial condition, large eddy simulation (LES) is used to calculate the unsteady flow field, the number of calculation steps is 1000 steps, the time step is 0.0005s, and each step is iterated 10 times; the control equation of the large eddy simulation (LES) is:

[0064] ;

[0065] Where: is the x-component of the gas flow velocity after filtering, is the y-component of the filtered gas flow velocity, and i≠j=0,1,2; t is time; x j is the y-component of the vortex scale; is the fluid density; is the pressure after filtering; x i is the x-component of the vortex scale; is the viscosity coefficient; is the sub-grid stress;

[0066] S23) After the flow field reaches dynamic equilibrium, the acoustic perturbation equation (APE) is used to calculate the acoustic field, with 500 calculation steps, a time step of 0.0001s, and 10 iterations per step; the acoustic perturbation equation (APE) is:

[0067] ;

[0068] Where: c is the speed of sound; is the sound pressure; t is the time; is the time average speed; is the fluid dynamic pressure; is the damping coefficient; is the time step; , is the local Courant number, which reflects the movement distance of the fluid in the grid, where is the grid spacing, Delta is the increment. , x n and x n-1are two adjacent nodes in the grid; in addition, is the nabla operator, which is a vector differential operator. The specific calculation method is as follows:

[0069]

[0070] Among them: I, J, K represent the direction components along the positive directions of the three coordinate axes x, y, and z respectively;

[0071] The boundary conditions for the above calculations are set as follows: the inlet is the velocity inlet, which is set to 111 m / s (equivalent to simulating a train running at 400 km / h); the outlet is the pressure outlet, which is set to 0 Pa; the ground is a non-slip wall, and the other surfaces are symmetrical boundaries, simulating the train running in an open-air environment;

[0072] S3) collects the sound pressure data at each grid node on several cross sections at different spatial heights in the pantograph area and parallel to the ground (this embodiment takes the collection of the sound pressure data at each grid node on the pantograph head cross section, the arm connection cross section and the pantograph cavity bottom surface as an example, such as Figure 4 As shown, but not limited to the three sections described in the embodiment, any section located at different spatial heights in the pantograph area and parallel to the ground is acceptable, and the number of sections is not limited. The more sections there are, the more detailed and accurate the simulation analysis results are); a total of 1500 steps are collected, and the time step is 0.0001s; the collected sound pressure data is stored in an EXCEL table, and the collected sound pressure data includes the number of each point, the three-dimensional coordinate value of each point, and the sound pressure value at each point;

[0073] S4) The collected sound pressure data is processed and analyzed using the spectral proper orthogonal decomposition (SPOD) method. The specific steps are as follows:

[0074] S41) importing the sound pressure data collected in step S3) into MATLAB software in the form of compressed snapshots;

[0075] S42) The snapshot matrix is ​​constructed by MATLAB software as Q=[q1,q2,…,q m ], where q m is an n-row, 1-column matrix containing the sound pressure information of each grid node on the cross section at time m, where n is the number of grid nodes on the cross section;

[0076] S43) Using the Welch method in MATLAB software, the snapshot matrix constructed in step S42) is divided into multiple overlapping block matrices to obtain a block matrix q with N0 adjacent overlapping snapshots. Nf , where Nf is the number of snapshots included in each block matrix;

[0077] S44) MATLAB software performs discrete Fourier transform and windowing on each block matrix, and then extracts the specified frequency f g The snapshots under are rearranged to obtain a new block matrix, and then the Fourier components of each block are arranged according to frequency to form a new snapshot matrix;

[0078] S45) The eigenvalue problem of each frequency-related matrix is ​​solved by MATLAB software, that is, the SPOD mode and the corresponding modal energy at each eigenfrequency are obtained, and then the modes are sorted according to the size of the modal energy to obtain the relationship diagram between the modal energy of each order and the frequency, thereby obtaining the noise modal energy distribution diagram at each section.

[0079] Figures 5 to 7 They are the noise modal energy distribution diagram of the bow head section, the noise modal energy distribution diagram of the arm connection section and the noise modal energy distribution diagram of the pantograph cavity bottom surface obtained after processing by MATLAB software. Each curve in the figure corresponds to the relationship curve between a certain order of modal energy and frequency, and they are all changing relationships from the first order to the Nth order from top to bottom.

[0080] Depend on Figures 5 to 7 It can be seen from the results shown that: the noise peak frequency of the bow head section is 1250Hz, and the distribution frequency range of the main noise energy of the bow head section is 700-1350Hz; the noise peak frequency of the cross section at the arm connection is 1015Hz, and the distribution frequency range of the main noise energy of the cross section at the arm connection is 550-1480Hz; the noise peak frequency of the bottom surface of the pantograph cavity is 935Hz, and the distribution frequency range of the main noise energy of the bottom surface of the pantograph cavity is 450-1420Hz; in addition, it can be seen from the results shown that as the spatial heights of the bow head section, the arm connection section, and the bottom surface of the pantograph cavity decrease successively, the distribution frequency range of the main noise energy is gradually widening.

[0081] From the above, it can be seen that the method of the present invention can be used to simulate and analyze the noise modal energy distribution diagram at each cross section at different spatial position heights in the pantograph area of ​​a high-speed train. From the obtained noise modal energy distribution diagram, the change of the peak frequency of each order modal energy and the distribution frequency range of the main energy can be known, and the change trend of the frequency range with the spatial position height can be known, so as to obtain the near-field sound pressure distribution and spatial propagation characteristics of the pantograph, thereby providing a theoretical basis for the near-field aerodynamic noise reduction of the pantograph of a high-speed train, which is of great significance and application value for solving the aerodynamic noise of high-speed trains.

[0082] Finally, it is necessary to point out that the above description is only used to further explain the technical solution of the present invention in detail and cannot be understood as limiting the scope of protection of the present invention. Some non-essential improvements and adjustments made by technicians in this field based on the above content of the present invention, such as changes in the noise area and changes in the position of the spatial cross-section, etc., all fall within the scope of protection of the present invention.

Claims

1. A method for simulating and analyzing the near-field noise modal energy distribution of a high-speed train pantograph, characterized in that: The steps include: S1) Establish a 3D finite element model of a high-speed train with a pantograph and perform meshing; S2) performing fluid simulation analysis on the grid model established in step S1); S3) collecting sound pressure data at each grid node on several cross sections located at different spatial heights in the pantograph area and parallel to the ground; S4) The collected sound pressure data are processed and analyzed using the spectral proper orthogonal decomposition (SPOD) method, that is, a noise modal energy distribution diagram between SPOD modal energy and frequency is obtained.

2. The method according to claim 1, characterized in that The step S1) includes the following specific operations: S11) Create a three-car train model with the first and last cars and the pantograph car; S12) Create a computational domain, including the air inlet and outlet, the bottom ground, and the symmetric boundaries of the remaining three surfaces; S13) removing the train model in the computational domain created in step S12) and leaving only the part where the air exists, and then performing meshing.

3. The method according to claim 2, characterized in that: Step S12) creates a computational domain in accordance with the national railway standard TB / T3503.4-2018 "Specifications for Numerical Simulation of Train Aerodynamic Performance", and step S13) uses the cutting volume provided by STAR-CCM+ for meshing.

4. The method according to claim 1, characterized in that: The step S2) includes the following specific operations: S21) adopts k- The turbulence model is used to calculate the steady flow field; S22) using the steady flow field calculated in step S21) as an initial condition, and performing unsteady flow field calculation using large eddy simulation (LES); S23) After the flow field reaches dynamic equilibrium, the acoustic perturbation equation (APE) is used to calculate the acoustic field.

5. The method according to claim 4, characterized in that Step S21) k- The governing equations of the turbulence model are: ； Where: is the fluid density; k is the turbulent kinetic energy; t is the time; x j is the y-component of the vortex scale; u j is the y-component of the gas flow velocity; is the x-component of the turbulent viscosity; is the turbulent viscosity; is the eddy viscosity model of Reynolds stress; S ij is the average velocity strain rate tensor; β* is the model constant, which is taken as 0.09; is the internal dissipation rate of turbulent kinetic energy; For the generated items; , ,β,F1, are model constants.

6. The method according to claim 4, characterized in that The governing equation of the large eddy simulation (LES) described in step S22) is: ； Where: is the x-component of the gas flow velocity after filtering, is the y-component of the filtered gas flow velocity, and i≠j=0,1,2; t is time; x j is the y-component of the vortex scale; is the fluid density; is the pressure after filtering; x i is the x-component of the vortex scale; is the viscosity coefficient; is the sub-grid stress.

7. The method according to claim 4, characterized in that The acoustic perturbation equation (APE) in step S23) is: ； Where: c is the speed of sound; is the sound pressure; t is the time; is the time average speed; is the fluid dynamic pressure; is the damping coefficient; is the time step; , is the local Courant number, which reflects the movement distance of the fluid in the grid, where is the grid spacing, Delta is the increment. x n and x n-1 are two adjacent nodes in the grid; in addition, is the nabla operator, which is a vector differential operator. The specific calculation method is as follows: Among them: I, J, K represent the direction components along the positive directions of the three coordinate axes x, y, and z respectively.

8. The method according to claim 1, characterized in that Step S2) The boundary conditions for fluid simulation analysis are set as follows: the inlet is a velocity inlet, set to 111 m / s; the outlet is a pressure outlet, set to 0 Pa; the ground is a no-slip wall, and the other surfaces are symmetrical boundaries, simulating the train running in an open-air environment.

9. The method according to claim 1, characterized in that: Step S3) A total of 1500 steps are collected, and the time step is 0.0001s.

10. The method according to claim 1, characterized in that The specific steps of processing and analyzing in step S4) are as follows: S41) importing the sound pressure data collected in step S3) into MATLAB software in the form of compressed snapshots; S42) Construct the snapshot matrix as Q=[q1,q2,…,q m ], where q m is an n-row, 1-column matrix containing the sound pressure information of each grid node on the cross section at time m, where n is the number of grid nodes on the cross section; S43) The snapshot matrix constructed in step S42) is divided into a plurality of overlapping block matrices by using the Welch method, and a block matrix q with N0 adjacent overlapping snapshots is obtained. Nf , where Nf is the number of snapshots included in each block matrix; S44) Perform discrete Fourier transform and windowing on each block matrix, and then extract the specified frequency f g The snapshots under are rearranged to obtain a new block matrix, and then the Fourier components of each block are arranged according to frequency to form a new snapshot matrix; S45) solve the eigenvalue problem of each frequency correlation matrix, that is, obtain the SPOD mode and the corresponding modal energy at each eigenfrequency, and then sort the modes according to the size of the modal energy to obtain the relationship diagram between the modal energy of each order and the frequency, so as to obtain the noise modal energy distribution diagram at each section.

Images