Running part framework fatigue damage assessment method based on motor vibration

Abstract

The invention discloses a motor vibration-based walking part framework fatigue damage assessment method, which comprises the following steps of: performing vibration test and dynamic stress test at a specified position of a bogie to obtain a motor acceleration signal and dynamic hot spot stress, and preprocessing the motor acceleration signal and the dynamic hot spot stress; a motor acceleration signal is used as input excitation, and the power spectrum density of the dynamic hot spot stress is calculated through a transfer function and the real-time motor acceleration; based on the obtained power spectrum density of the dynamic hot spot stress, reconstructing a stress time history by using fast Fourier inversion; and based on the stress time history obtained by reconstruction, performing fatigue stress estimation by adopting a frequency domain method. The fatigue damage degree of the bogie frame is evaluated through motor vibration, and harm caused by fatigue fracture of the frame is avoided.

Description

Technical Field

[0001] The invention relates to the field of train flaw detection, and in particular to a method for evaluating fatigue damage of a running gear frame based on motor vibration. Background Art

[0002] As an important part of rail vehicles, the running gear frame bears complex loads and frequent starting and braking. Due to the changing line conditions, fatigue failure of rail vehicles is very common, and real-time monitoring of abnormal vibration of rail vehicles is very necessary.

[0003] Therefore, the structural safety monitoring system in the railway industry has received more and more attention. The extensive application of on-board sensors and roadside equipment helps to estimate the damage of key components and monitor wheel-rail deterioration.

[0004] Generally speaking, the existing structural safety monitoring methods in the railway industry mainly use acceleration signals, inertial system signals, current signals and measured strain, among which the main indicators of comfort and safety are mostly obtained through indirect measurement. The above abnormal vibration monitoring methods focus on the identification of wheel-rail irregularities or the comfort assessment of the vehicle body. There are few research reports on the resonant vibration monitoring and damage assessment of bogie frames.

[0005] In general, the structural health monitoring methods of bogie frames lack diversification and need further research. Due to the complex external loads, more effective, simpler and less computationally intensive methods are needed for abnormal vibration monitoring and damage estimation of bogie frames. Due to space limitations and experimental conditions, the number of available sensors is relatively small. The main modes of the bogie frame that are sensitive to cracks can usually be determined by test results or finite element analysis. Therefore, due to certain failure modes of the bogie frame, the number of sensor channels used can be reduced. Therefore, it is meaningful to use finite sensors for fatigue damage monitoring of subway bogies. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for evaluating fatigue damage of a running gear frame based on motor vibration.

[0007] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is: A method for evaluating fatigue damage of a running gear frame based on motor vibration comprises the following steps: S1. Perform vibration test and dynamic stress test at the designated position of the bogie, obtain motor acceleration signal and dynamic hot spot stress and perform preprocessing; S2, using the motor acceleration signal as input excitation, and calculating the power spectrum density of the dynamic hot spot stress through the transfer function and the real-time motor acceleration; S3, based on the obtained power spectrum density of the dynamic hot spot stress, reconstruct the stress time history using inverse fast Fourier transform; S4. Based on the reconstructed stress time history, the frequency domain method is used to estimate fatigue stress.

[0008] Furthermore, the specific method of preprocessing in S1 is: S11, dividing the acquired motor acceleration signal and dynamic hot spot stress according to the same window width to calculate the running mean and variance of the motor acceleration signal and the dynamic hot spot stress in each window; S12, determining the confidence interval of the number of runs based on the obtained running mean and variance, if the number of runs falls outside the confidence interval, the signal is considered to be a non-stationary signal, and calculating its non-stationary index; S13. Calculate the running test value of the dynamic hot spot stress, and determine the non-stationarity of the calculated running test value based on the non-stationarity index. Furthermore, the calculation method of the running mean and variance in S11 is:

[0009]

[0010] In the formula, is the running mean, is the total number of windows, is the number of the ith observation, is the variance, is the number of runs.

[0011] Furthermore, the confidence interval in S12 is defined as:

[0012] In the formula, is the running mean, is the standard deviation, is the weight.

[0013] Furthermore, the calculation method of the non-stationary index in S12 is:

[0014] In the formula, is the non-stationary index, is the number of runs, is the running mean.

[0015] Furthermore, the power spectrum density of the dynamic hot spot stress in S2 is calculated as follows:

[0016] In the formula, is the transfer function between the motor vertical acceleration and the point stress, is the transfer function The complex conjugate of .

[0017] Furthermore, according to the motor vibration acceleration spectrum, combined with the frequency domain transfer function of motor vibration and frame stress, the frame stress spectrum is obtained, and the stress time mileage curve is obtained by Fourier inverse transform. The Fourier transform of the time history curve can be expressed as:

[0018] in, is a frequency domain function, is a time domain function, is an exponential function, is the angular frequency.

[0019] Furthermore, the specific method of using the frequency domain method to perform fatigue stress in S4 is:

[0020] In the formula, is the fatigue damage strength of the rain flow counting method, is the peak frequency, is the fatigue strength of the SN curve, k is the slope of the SN curve, is the probability density function of the stress amplitude S of the rainflow counting method.

[0021] The present invention has the following beneficial effects: The fatigue damage degree of the bogie frame is assessed through motor vibration to avoid the harm caused by fatigue fracture of the frame. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The figure is a flow chart of the method for assessing fatigue damage of a running gear frame based on motor vibration of the present invention.

[0023] Figure 2 Schematic diagram of the transfer rate of dynamic stress at different frequencies according to an embodiment of the present invention.

[0024] Figure 3a A schematic diagram comparing the first segment predicted and measured stress power spectrum densities selected for an embodiment of the present invention.

[0025] Figure 3b A schematic diagram comparing the second segment predicted and measured stress power spectrum densities selected for an embodiment of the present invention.

[0026] Figure 4a The figure is a schematic diagram for comparing the predicted and measured y-axis stress power spectrum density at 100 MPa root mean square value according to an embodiment of the present invention.

[0027] Figure 4b The figure is a schematic diagram comparing the predicted and measured y-axis stress power spectrum density at 80 MPa root mean square value according to an embodiment of the present invention.

[0028] Figure 4c The figure is a schematic diagram comparing the cross-sectional non-uniformity index of the predicted and measured y-axis stress power spectrum density at a root mean square value of 100 MPa according to an embodiment of the present invention.

[0029] Figure 4d It is a schematic diagram comparing the cross-sectional non-uniformity index predicted and measured by the y-axis stress power spectrum density at a root mean square value of 80 MPa according to an embodiment of the present invention.

[0030] Figure 5a This is a schematic diagram of the probability density distribution function estimation result calculated by the Dirlik method for the time period of the first irregularity index in an embodiment of the present invention.

[0031] Figure 5b This is a schematic diagram of the estimation result of the probability density distribution function calculated by the Dirlik method for the time period of the second irregularity index in an embodiment of the present invention.

[0032] Figure 5c This is a schematic diagram of the estimation result of the probability density distribution function calculated by the Dirlik method for the time period of the third irregularity index in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0034] A fatigue damage assessment method for running gear frame based on motor vibration, such as Figure 1 As shown, the following steps are included: S1. Perform vibration test and dynamic stress test at the designated position of the bogie, obtain motor acceleration signal and dynamic hot spot stress and perform preprocessing; In this embodiment, actual measurements are performed on the bogie frame: sensors are installed at designated locations, vibration tests and dynamic stress tests are performed, and the required acceleration signals and dynamic hot spot stresses are obtained.

[0035] The hot spot stress location was determined by field testing. Only the vertical motor acceleration was used as the input excitation of the bogie frame, which was measured in real time by the motor's sensor. The y-axis stress at the bogie frame crossbeam was the only output. The entire system can then be considered a single-input single-output system, and the power spectral density stress for:

[0036] In the formula is the transfer function between the motor vertical acceleration and the point dynamic stress. is the transfer function In the above transfer function estimation method, the cross-correlation power spectral density acceleration of the input and the power spectral density acceleration of the output stress are used.

[0037] Due to the frequent starting and braking of subway vehicles during operation, the measured stresses and accelerations tend to be non-stationary between the two stations. To quantify the non-stationarity of the signal, the vibration signal is segmented by equal window widths using non-stationarity indicators. Then, the changes in the statistical properties relative to the entire signal are analyzed in each window. The runs are defined based on the same observation sequence, which are classified into one of two mutually exclusive names, such as greater or less than the RMS value of the entire signal. Assumptions is the number of observations in i, and N is the total number of windows. Then the running mean and variance as follows:

[0038] Therefore, the distribution of the number of runs r is determined by the confidence interval:

[0039] in Take 1.96, considering the 95% confidence level. If the number of runs falls outside the confidence interval, the signal is considered non-stationary. Therefore, the non-stationarity index Determined by the number of runs divided by the expected average number of runs:

[0040] Run test value Defined as

[0041] The typical stress time history is processed to determine the non-stationarity. The thick red line represents the standard deviation of each window, and the dotted blue line represents and The sign of the running test value depends on whether the standard deviation of each window lies within the bounds. For example, the non-stationarity index = 0.1933, which is much smaller than 1, indicating a non-stationary process. Therefore, the transfer function of the vertical motor acceleration to the bogie frame dynamic stress was calculated at a shorter time interval, where the measured signal tends to be stationary.

[0042] S2, using the motor acceleration signal as input excitation, and calculating the power spectrum density of the dynamic hot spot stress through the transfer function and the real-time motor acceleration; Estimation of frequency domain dynamic stress spectrum through frequency domain input excitation and transfer function: For a single degree of freedom system, consider a stationary random vibration x (t) acting on a linear mechanical system and its response y (t). Their Fourier transform is:

[0043] in is the transfer function matrix. Then we square both sides of the equation and define it according to the self-power spectrum to get

[0044]

[0045] in and It is the power spectral density of the input and output responses For a single-input / single-output (SISO) system, the transfer function is estimated as:

[0046] in are the x and y coordinates of the cross-correlation power spectral density function, is the x-coordinate of the function of the power spectral density. For a MIMO system, the mathematical form is the same as for a SIGOUT system, where and They represent the cross-correlation power spectral density matrix and power spectral density matrix of multiple-input and multiple-output respectively.

[0047] The vertical acceleration of the motor can be used as input excitation and acquired through vehicle-mounted equipment. Then, the power spectral density of the hot spot stress as the system output is calculated through the transfer function and the real-time motor acceleration.

[0048] Due to the different orders of wheel polygons, the peak of the autospectral density amplitude appears in the range of 60Hz to 80Hz. The results show that the 5th to 9th order wheel polygons are dominant. When the running speed is 50km / h to 80km / h, the corresponding excitation frequency is 42Hz to 67Hz. The polygon of the right wheel is the most serious, reaching 0.8mm, and the corresponding autospectral density amplitude peak is also large. The peak of the cross-correlation power spectral density amplitude of the motor mainly appears in the 50Hz and 74Hz frequency bands, which is significantly larger than other frequency bands. The results show that in the 40Hz to 74Hz frequency band, especially in the 50Hz frequency band, the y-axis stress is highly correlated with the vertical acceleration of the motor.

[0049] S3, based on the obtained power spectrum density of the dynamic hot spot stress, reconstruct the stress time history using inverse fast Fourier transform; Then, using the autospectral density of the vertical acceleration of the shaft axis and the cross-correlation power spectral density of the dynamic stress to acceleration mentioned above, the transfer function can be calculated in a short time interval. The input-to-output transfer function shows a high transfer rate from the motor vertex acceleration to the dynamic stress at a frequency close to 50Hz. In addition, the transfer rate is significantly greater in the frequency bands of 30Hz to 55Hz and below 10Hz. From the experimental results, only the high-frequency components above 20Hz are obvious, such as Figure 2 As shown. It can also be seen that Figure 2 During the whole operation, the frequency around 8 Hz exists. It can be inferred from the wheel rotation frequency (80 km / h running speed f = v / 2xR = 8.1 Hz). Due to the nonlinearity and uncertainty of the whole system, the calculated average transfer function is used.

[0050] After obtaining the power spectral density of the relevant stress, the stress time history is reconstructed using the inverse fast Fourier transform (IFFT). Since there is no phase information in the power spectral density of the stress, the reconstruction of the stress time history is based on the random phase uniformly distributed at different frequencies. Due to the randomness of the phase, the reconstructed time series is a Gaussian quasi-random sequence. However, on rail vehicles, the dynamic stress of the bogie frame appears as a non-Gaussian dynamic stress over a large time span (tens of seconds). An underestimation of the stress kurtosis will lead to a lower estimate of the structural damage. Therefore, we will show the time-varying kurtosis over a small time span, and the root mean square value of the signal can be obtained in both the time domain and the frequency domain.

[0051] S4. Based on the reconstructed stress time history, the frequency domain method is used to estimate fatigue stress.

[0052] In the frequency domain method, the autocorrelation function and the power spectral density function are used to describe the characteristics of the stationary random process in the time domain and frequency domain respectively. According to the Wiener-Sinchin theorem, the autocorrelation function and the power spectral density function form the following Fourier transform:

[0053]

[0054] when The interval is located When , the one-sided power spectral density function is defined as

[0055] Similar to the definition of statistical moments, the spectral moment set used to describe statistical properties is given

[0056] Bandwidth parameter or irregularity indexr for

[0057] The irregularity index r ranges from 0 to 1. Another form is the van Mark spectral parameter , if x(t) is narrowband, then r approaches 1, and if it is wideband, then r converges to zero. The average square root of x(t) is , average exceedance rate and its peak frequency is They are ,

[0058] The probability density function of the stress amplitude S of the rainflow counting method is calculated as follows:

[0059]

[0060] in is the normalized stress amplitude, is the average frequency, The spectrum parameters are

[0061]

[0062] Rainflow counting fatigue damage strength is the damage per unit time under the Palmgren Miner law,

[0063] Where C and k are the fatigue strength and slope of the SN curve, respectively.

[0064] Based on the obtained transfer function and the power spectral density of the vertical acceleration of the click, the y-axis stress power spectral density of the relevant points of the bogie frame in each time interval can be calculated. As shown in Figure 3, the predicted and measured stress power spectral densities are compared with two selected sections with a time interval of 2s. The peaks of the stress power spectral density usually appear around 4Hz and 50Hz. It can be seen that the predicted results are quite comparable with the measured results. The predicted results tend to overestimate the amplitude of the power spectral density results in the frequency range above 100Hz. In this case, the nonlinearity may come from the frequency variation characteristics of the main suspension composed of rubber springs. The RMS values ​​and inhomogeneity indicators of the predicted and measured y-axis stress power spectral densities are compared to verify the effectiveness of the method. As shown in Figure 4 (a) and (b), the RMS values ​​of the 100 and 80 sections are plotted, and the predicted results are in good agreement with the measured values. The inhomogeneity indicators of the corresponding sections, as shown in Figure 4 (c) and (d), also prove the effectiveness of the stress power spectral density prediction. The irregularity index describes the shape of the signal power spectral density, and higher values ​​close to 1 represent a narrowband process.

[0065] The predicted irregularity index is mainly in the range of 0.6 to 0.8, while the measured irregularity index varies in the range of 0.7 to 0.9, which may be caused by the non-Gaussian nature of the process. Overall, the proposed method using limited sensors has been shown to be effective in predicting the y-axis stress power spectral density in terms of the amplitude and bandwidth of the dominant frequencies.

[0066] Then, the stress power spectrum density is used to estimate the probability density distribution function of the stress amplitude at each point in each time interval. The probability density distribution function of the stress amplitude is calculated using the measured and predicted power spectrum density and compared with the rainflow counting results. Figure 5a As shown in a, b and c, three time periods with different irregularity indices are plotted. The estimated results of the probability density distribution function calculated by the Dirlik method are consistent with the rainflow counting results. With the increase of the non-uniformity index (0.69), the proportion of large stress amplitudes in the range of 10MPa to 40MPa increases. The irregularity index nearby is low, at 0.5, and the proportion is greatly reduced to close to zero, and the dominant stress amplitude is within 10MPa. The predicted power spectral density is basically consistent with the calculated results of the measured power spectral density, and the error is within an acceptable range. The damage value is then calculated using the frequency domain method.

[0067] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0068] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0069] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0070] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0071] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A method for assessing fatigue damage of a running gear frame based on motor vibration, characterized in that: The steps include: S1. Perform vibration test and dynamic stress test at the designated position of the bogie, obtain motor acceleration signal and dynamic hot spot stress and perform preprocessing; S2, using the motor acceleration signal as input excitation, and calculating the power spectrum density of the dynamic hot spot stress through the transfer function and the real-time motor acceleration; S3, based on the obtained power spectrum density of the dynamic hot spot stress, reconstruct the stress time history using inverse fast Fourier transform; S4. Based on the reconstructed stress time history, the frequency domain method is used to estimate fatigue stress.

2. The method for evaluating fatigue damage of a running gear frame based on motor vibration according to claim 1 is characterized in that: The specific method of preprocessing in S1 is: S11, dividing the acquired motor acceleration signal and dynamic hot spot stress according to the same window width to calculate the running mean and variance of the motor acceleration signal and the dynamic hot spot stress in each window; S12, determining the confidence interval of the number of runs based on the obtained running mean and variance, if the number of runs falls outside the confidence interval, the signal is considered to be a non-stationary signal, and calculating its non-stationary index; S13. Calculate the running test value of the dynamic hot spot stress, and determine the non-stationarity of the calculated running test value based on the non-stationarity index.

3. The method for evaluating fatigue damage of a running gear frame based on motor vibration according to claim 2 is characterized in that: The calculation method of the running mean and variance in S11 is: In the formula, is the running mean, is the total number of windows, is the number of the i-th observation, is the variance, is the number of runs.

4. The method for evaluating fatigue damage of a running gear frame based on motor vibration according to claim 2 is characterized in that: The confidence interval in S12 is defined as: In the formula, is the running mean, is the standard deviation, is the weight.

5. The method for evaluating fatigue damage of a running gear frame based on motor vibration according to claim 2 is characterized in that: The calculation method of the non-stationary index in S12 is: In the formula, is the non-stationary index, is the number of runs, is the running mean.

6. The method for evaluating fatigue damage of a running gear frame based on motor vibration according to claim 1, characterized in that: The power spectrum density of the dynamic hot spot stress in S2 is calculated as follows: In the formula, is the transfer function between the motor vertical acceleration and the point stress, is the transfer function The complex conjugate of .

7. The method for evaluating fatigue damage of a running gear frame based on motor vibration according to claim 1, characterized in that: According to the motor vibration acceleration spectrum, combined with the frequency domain transfer function of motor vibration and frame stress, the frame stress spectrum is obtained, and the stress time mileage curve is obtained by Fourier inverse transform. The Fourier transform of the time history curve can be expressed as: in, is a frequency domain function, is a time domain function, is an exponential function, is the angular frequency.

8. The method for evaluating fatigue damage of a running gear frame based on motor vibration according to claim 1, characterized in that: The specific method of using the frequency domain method to calculate fatigue stress in S4 is: In the formula, is the fatigue damage strength of the rain flow counting method, is the peak frequency, is the fatigue strength of the SN curve, k is the slope of the SN curve, is the probability density function of the stress amplitude S of the rainflow counting method.

Images