Vibration fatigue life evaluation method for unsprung motor controller

Abstract

The invention relates to the field of motor controller vibration reliability, in particular to a vibration fatigue life evaluation method for an unsprung motor controller, and the method comprises the steps: collecting vibration signals of an electric drive axle (integrating a motor, a motor controller and a main speed reducer) in a typical use scene of a vehicle; and establishing a standard spectrum suitable for the unsprung motor controller according to the vibration signal, and performing random vibration intensity simulation, fatigue life evaluation and random vibration fatigue test on the unsprung motor controller random vibration analysis finite element model by using the standard spectrum, so that the vibration fatigue life evaluation of the unsprung motor controller can be realized. The test verification frequency of a field scene can be reduced, the development cost is saved, and the vibration durability development process of the unsprung motor controller is accelerated.

Description

Technical Field

[0001] The present invention relates to the field of vibration reliability of motor controllers, and in particular to a vibration fatigue life assessment method for an unsprung motor controller. Background Art

[0002] In micro-new energy commercial vehicles, the motor controller is the core component responsible for converting the direct current transmitted from the battery into three-phase alternating current, controlling the motor torque and speed, and ensuring the normal operation of the entire vehicle. Traditional motor controllers are mainly installed on the frame above the chassis of micro-new energy commercial vehicles using a sprung installation method, but the sprung installation method often takes up a large amount of space in micro-new energy commercial vehicles and increases the use of high-voltage wiring harnesses and connectors. In order to optimize the overall layout space, reduce costs, and increase the endurance of the entire vehicle, micro-new energy commercial vehicles currently mainly use an unsprung installation method, by integrating the motor controller into the motor and installing it together with the motor on the drive axle to form a three-in-one electric drive axle. This motor controller set on the three-in-one electric drive axle is usually called an "unsprung motor controller."

[0003] The unsprung motor controller has no leaf spring to reduce the vibration excitation from the road, and is closer to the motor, so it will subject the motor controller to greater vibration (such as Figure 1 As shown). However, most electronic components in the motor controller have poor rigidity (such as potentiometers, relays, etc.), and are more susceptible to vibration excitation from the ground, motors, and reducers, making them more susceptible to vibration fatigue failure, posing a huge threat to the driver's life and property safety. In order to ensure that the motor controller does not fail due to vibration fatigue during vehicle driving, it is necessary to carry out a vibration fatigue life assessment during the development stage of the motor controller. However, at this stage, there is no clear industry standard or national standard for the vibration durability standard and definition of unsprung motor controllers. The micro-new energy commercial vehicle industry can only refer to the national standard GB / T28406.3 for passenger car non-elastic body vibration PSD (Power spectral density) for vibration durability fatigue life assessment. This standard is mainly for fuel vehicle electronic and electrical equipment, and has low reference value for unsprung motor controllers of micro-new energy commercial vehicles.

[0004] Therefore, how to effectively evaluate the vibration fatigue life of the unsprung motor controller for micro new energy commercial vehicles has always been an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0005] The purpose of the present invention is to address the corresponding deficiencies in the existing technology and provide a vibration fatigue life assessment method for an unsprung motor controller. By collecting vibration signals of an electric drive axle (integrating a motor and a motor controller) under typical vehicle usage scenarios, and establishing a standard spectrum suitable for the unsprung motor controller based on the vibration signals, the standard spectrum is used to perform random vibration intensity simulation, fatigue life assessment, and random vibration fatigue testing on a finite element model of random vibration analysis of the unsprung motor controller, thereby achieving effective assessment of the vibration fatigue life of the unsprung motor controller.

[0006] The purpose of the present invention is to adopt the following scheme to achieve:

[0007] A vibration fatigue life assessment method for an unsprung motor controller includes the following steps:

[0008] 1) Design and manufacture vibration tooling for vibration fatigue life assessment based on the motor controller and the corresponding motor;

[0009] 2) Before vibration fatigue life assessment, corresponding finite element models are established for the vibration fixture, motor, and motor controller. These finite element models are then combined into a finite element model for random vibration analysis of the unsprung motor controller.

[0010] 3) When performing vibration fatigue life assessment, several vibration measurement points are set on the electric drive axle, and corresponding three-axis acceleration sensors are set at each vibration measurement point to obtain several vibration spectra;

[0011] 4) Using all vibration spectra, establish a standard spectrum;

[0012] 5) Input the canonical spectrum as the simulated load into the finite element model for random vibration analysis of the unsprung motor controller to perform random vibration intensity simulation. Based on the simulation results, the motor controller is structurally optimized until the strength of the motor controller meets the requirements.

[0013] 6) Perform a swept-frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain the stress-frequency response relationship. This stress-frequency response relationship is then imported into fatigue analysis software. The canonical spectrum is then input as the load into the fatigue analysis software to calculate the fatigue life of the unsprung motor controller.

[0014] Preferably, in step 4), the specific method of establishing a standard spectrum using all vibration spectra includes:

[0015] 4-1) Calculate the power spectrum density of each vibration spectrum and merge each power spectrum density into a feature sample;

[0016] 4-2) Set a fitting curve and use the fitting curve to normalize the characteristic samples according to the following formula to obtain normalized characteristic samples;

[0017] 4-3) Calculating the mean, variance, and upper tolerance coefficient of the normalized characteristic sample, and calculating the upper limit of the normalized characteristic sample based on the mean, variance, and upper tolerance coefficient;

[0018] 4-4) Calculate the measured spectrum based on the upper limit of the normalized characteristic sample;

[0019] 4-5) According to the frequency domain distribution characteristics of the measured spectrum, the measured spectrum is divided into frequency bands, and the measured spectrum after frequency band division is converted into a standard spectrum, and the loading time of the standard spectrum is compressed.

[0020] Preferably, in step 4-1), the specific method of calculating the power spectrum density of each vibration spectrum and merging each power spectrum density into a feature sample includes:

[0021] 4-1-1) Calculate the power spectrum density of each vibration spectrum, and calculate the root mean square of each power spectrum density, as well as the mean and variance of each root mean square;

[0022] 4-1-2) Based on the means and variances calculated in step 4-1-1), calculate the F statistic and t statistic of the power spectral density corresponding to each mean and variance;

[0023] 4-1-3) Perform F-test and t-test on the F-statistic and t-statistic calculated in step 4-1-2), and filter each power spectrum density in the following way to obtain several power spectrum densities that meet the merging conditions:

[0024] If the F statistic passes the F test and the t statistic passes the t test, the corresponding power spectrum density is retained;

[0025] If the F statistic fails the F test, or the t statistic fails the t test, the corresponding power spectrum density is discarded;

[0026] 4-1-4) The power spectrum density obtained in step 4-1-3) is merged to obtain feature samples.

[0027] Preferably, the specific formulas of the F test and t test are:

[0028]

[0029] Where, F (M-1,M-1)；α / 2 is the α / 2 quantile of the F distribution with degrees of freedom (M-1,M-1), F(i,m) is the F statistic, F (M-1,M-1)；(1-α / 2)is the (1-α / 2) quantile of the F distribution with degrees of freedom (M-1,M-1), t(i,m) is the t statistic, t 2(M-1)；(1-α / 2) is a t-distribution (1-α / 2) with 2(M-1) degrees of freedom.

[0030] Preferably, the specific method of performing random vibration intensity simulation on the finite element model of random vibration analysis of the unsprung motor controller includes:

[0031] 5-1) Use tensile testing to obtain the tensile strength limit of the motor controller material;

[0032] 5-2) Using the canonical spectrum as the simulation load, input it into the finite element model for random vibration analysis of the unsprung motor controller, perform random vibration intensity simulation on the finite element model for random vibration analysis of the unsprung motor controller, and obtain the root mean square stress value;

[0033] 5-3) Based on the RMS stress value and the material tensile strength, use the following formula to determine whether the motor controller's vibration intensity is qualified:

[0034] If 3×root mean square stress value is less than the tensile limit of the material, the vibration strength of the motor controller is qualified;

[0035] If the 3×root mean square stress value is ≥ the material tensile limit, the vibration strength of the motor controller is unqualified and the structural design of the motor controller is optimized.

[0036] Preferably, the specific method of performing random vibration fatigue life assessment on the random vibration analysis finite element model of the unsprung motor controller includes:

[0037] 6-1) Set the amplitude and frequency range of the swept frequency vibration simulation;

[0038] 6-2) Perform a swept frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain the stress-frequency response results of the unsprung motor controller;

[0039] 6-3) Importing the swept frequency vibration stress-frequency response results obtained in step 5-2) into fatigue analysis software, and using the canonical spectrum as the load input and the canonical spectrum time as the loading time;

[0040] 6-4) Estimate the material's stress-life curve using fatigue analysis software based on the material's tensile strength. Calculate the fatigue life of the unsprung motor controller based on the stress-life curve.

[0041] Preferably, a random vibration fatigue bench test is also involved to further verify the results of the random vibration fatigue life analysis of the unsprung motor controller.

[0042] Preferably, the plurality of vibration measuring points include a vibration measuring point at the connection between the motor and the main reducer, a vibration measuring point at the left and right leaf spring seats, and a vibration measuring point at the motor connection bracket.

[0043] The beneficial effects of the present invention are as follows:

[0044] A vibration fatigue life assessment method for an unsprung motor controller includes the following steps:

[0045] 1) Design and manufacture vibration tooling for vibration fatigue life assessment based on the motor controller and the corresponding motor;

[0046] 2) Before the vibration fatigue life assessment, finite element models of the motor, motor controller, and vibration tooling are established for the vibration tooling, motor, and motor controller involved in the vibration fatigue life assessment. The finite element models of the motor, motor controller, and vibration tooling are then combined into a finite element model for random vibration analysis of the unsprung motor controller.

[0047] 3) When performing vibration fatigue life assessment, several vibration measurement points are set on the electric drive axle that integrates the motor and motor controller, and a corresponding three-axis acceleration sensor is set at each vibration measurement point. The vibration signals of each vibration measurement point are collected by each three-axis acceleration sensor under typical vehicle usage scenarios and stored as corresponding vibration spectra;

[0048] 4) Using all vibration spectra, establish a standard spectrum;

[0049] 5) Inputting the canonical spectrum as the simulated load into the finite element model for random vibration analysis of the unsprung motor controller, performing random vibration intensity simulation on the finite element model for random vibration analysis of the unsprung motor controller, and optimizing the motor controller structure based on the simulation results until the strength of the motor controller meets the requirements;

[0050] 6) Perform a swept-frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain the stress-frequency response relationship. This stress-frequency response relationship is imported into fatigue analysis software, and the canonical spectrum is input as the load into the fatigue analysis software. The fatigue life of the unsprung motor controller is then calculated based on the stress-frequency response relationship.

[0051] The present invention constructs a standard spectrum based on the vibration signals of the electric drive axle (integrated motor and motor controller) collected under typical vehicle usage scenarios, and uses the standard spectrum for random vibration intensity simulation experiments and swept frequency vibration simulation experiments to obtain the theoretical value of the fatigue life of the unsprung motor controller. This can quickly and specifically evaluate the vibration fatigue life of the unsprung motor controller.

[0052] Preferably, a random vibration fatigue bench test is also involved to obtain a measured value of the fatigue life of the unsprung motor controller, so as to further verify the accuracy of the theoretical value of the fatigue life of the unsprung motor controller.

[0053] Preferably, the specific method of performing random vibration intensity simulation on the finite element model of random vibration analysis of the unsprung motor controller includes:

[0054] 5-1) Use tensile testing to obtain the tensile strength limit of the motor controller material;

[0055] 5-2) Using the canonical spectrum as the simulation load, input it into the finite element model for random vibration analysis of the unsprung motor controller, perform random vibration intensity simulation on the finite element model for random vibration analysis of the unsprung motor controller, and obtain the root mean square stress value;

[0056] 5-3) Based on the RMS stress value and the material tensile strength, use the following formula to determine whether the motor controller's vibration intensity is qualified:

[0057] If 3×root mean square stress value is less than the tensile limit of the material, the vibration strength of the motor controller is qualified;

[0058] If the 3×root mean square stress value is ≥ the material tensile limit, the vibration strength of the motor controller is unqualified and the structural design of the motor controller is optimized.

[0059] By performing random vibration intensity simulation on the finite element model of random vibration analysis of the unsprung motor controller, the present invention can comprehensively cover various complex factors that may arise in the actual operation of the motor controller, and provide detailed and high-precision analysis results, which helps to accurately identify potential weak links in the motor controller, thereby targetedly optimizing the structure and improving the design of the motor controller, which helps to improve the vibration intensity and stability of the motor controller.

[0060] Preferably, the plurality of vibration measuring points include a vibration measuring point at the connection between the motor and the main reducer, a vibration measuring point at the left and right leaf spring seats, and a vibration measuring point at the motor connection bracket.

[0061] By setting vibration measurement points at the connection between the motor and the final reducer, the left and right leaf spring seats, the motor connection bracket, and other places, the present invention can directly obtain vibration information of the motor and related components when the vehicle is running, so as to accurately reflect the actual vibration environment in which the unsprung motor controller is located, and provide detailed data support for accurately evaluating the vibration fatigue life of the unsprung motor controller.

[0062] The advantages of the present invention are as follows:

[0063] ① The present invention uses a combination of simulation experiments and field scenario tests to evaluate the vibration fatigue life of unsprung motor controllers. This can reduce the number of field scenario test verifications, save development costs, and accelerate the vibration durability performance development process of unsprung motor controllers.

[0064] ② The present invention can effectively identify and control the risk of vibration failure of the unsprung motor controller by conducting targeted vibration fatigue life analysis on the unsprung motor controller, which helps to discover potential fatigue weaknesses in the design stage of the electronic controller and make timely optimization and improvements to avoid product failures and safety hazards caused by fatigue failure.

[0065] Glossary

[0066] Vibration excitation refers to external forces that cause an object to vibrate. These forces can be changes in physical quantities such as force, displacement, velocity, or acceleration. The effects of vibration excitation refer to the various effects these external forces have on the vibration characteristics of an object and the performance of related systems.

[0067] Vibration fatigue failure: refers to the failure phenomenon caused by cumulative damage to materials, parts or structures under cyclic vibration loads, that is, cracks appear in the object and gradually expand, eventually leading to fracture or loss of its original function.

[0068] A three-axis accelerometer is a sensor that can simultaneously measure acceleration in three mutually perpendicular directions (typically the x, y, and z axes). It senses acceleration based on various physical principles, such as the inertial force generated by a mass under acceleration. Acceleration is measured by detecting changes in the mass's displacement, stress, or strain. This allows for precise measurement of changes in an object's motion in three dimensions, including linear and angular acceleration.

[0069] The F statistic is a commonly used statistic in analysis of variance (ANOVA) and some regression analyses. It is used to test whether two population variances are equal or whether multiple population means are equal. In the context of ANOVA, it measures the ratio of the between-group variation to the within-group variation.

[0070] t statistic: Mainly used in small sample cases to test whether the population mean is equal to a given value (single-sample t test), or to test whether two population means are equal (two-sample t test).

[0071] Random vibration fatigue testing: This method simulates random vibration conditions in a real-world environment and is used to evaluate the fatigue resistance of products or structures under random vibration loads. During the test, a specialized vibration test bench is used. A controller inputs irregular random input signals, subjecting the specimen to simulated random vibrations in three mutually perpendicular directions (usually the x, y, and z axes). BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 Schematic diagram of vibration signals of sprung and unsprung motor controllers;

[0073] Figure 2 Flowchart of the present invention;

[0074] Figure 3 Schematic diagram of a finite element model for random vibration analysis of an unsprung motor controller according to an embodiment of the present invention;

[0075] Figure 4 This is a schematic diagram of a random vibration test site in an embodiment of the present invention;

[0076] Figure 5 This is a flow chart of vibration fatigue life assessment of an unsprung motor controller in an embodiment of the present invention. DETAILED DESCRIPTION

[0077] like Figures 2 to 5 As shown, the vibration fatigue life evaluation method for the unsprung motor controller includes the following steps:

[0078] 1) Design and manufacture vibration tooling for vibration fatigue life assessment based on the motor controller and the corresponding motor;

[0079] 2) Before the vibration fatigue life assessment, finite element models of the motor, motor controller, and vibration tooling are established for the vibration tooling, motor, and motor controller involved in the vibration fatigue life assessment. The finite element models of the motor, motor controller, and vibration tooling are then combined into a finite element model for random vibration analysis of the unsprung motor controller.

[0080] 3) When performing vibration fatigue life assessment, several vibration measurement points are set on the electric drive axle that integrates the motor and motor controller, and a corresponding three-axis acceleration sensor is set at each vibration measurement point. The vibration signals of each vibration measurement point are collected by each three-axis acceleration sensor under typical vehicle usage scenarios and stored as corresponding vibration spectra;

[0081] 4) Using all vibration spectra, establish a standard spectrum;

[0082] 5) Inputting the canonical spectrum as the simulated load into the finite element model for random vibration analysis of the unsprung motor controller, performing random vibration intensity simulation on the finite element model for random vibration analysis of the unsprung motor controller, and optimizing the motor controller structure based on the simulation results until the strength of the motor controller meets the requirements;

[0083] 6) Perform a swept-frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain the stress-frequency response relationship. This stress-frequency response relationship is imported into fatigue analysis software, and the canonical spectrum is input as the load into the fatigue analysis software. The fatigue life of the unsprung motor controller is then calculated based on the stress-frequency response relationship.

[0084] Using the above method, the following examples are made:

[0085] 1) Design and manufacture vibration tooling for vibration fatigue life assessment based on the motor controller and corresponding motor that need vibration fatigue life assessment;

[0086] 2) Before the vibration fatigue life assessment, finite element models of the motor, motor controller, and vibration tooling are established for the vibration tooling, motor, and motor controller involved in the vibration fatigue life assessment. The finite element models of the motor, motor controller, and vibration tooling are then combined into a finite element model for random vibration analysis of the unsprung motor controller. The specific method is as follows:

[0087] 2-1) Create finite element models of the motor and motor controller in finite element software, perform modal simulation of the motor-motor controller two-in-one system, and record the simulation results;

[0088] 2-2) Conduct field-scenario modal experiments on the motor and motor controller of a certain type of micro-new energy commercial vehicle, and record the experimental results of the field scenario;

[0089] 2-3) Compare the simulation results with the experimental results of the actual scene, and modify the finite element model of the motor and motor controller based on the modal differences;

[0090] 2-4) Establish a finite element model of the vibration fixture, and equip the finite element model of the motor-motor controller two-in-one system to the finite element model of the vibration fixture, thereby completing the establishment of the finite element model of the random vibration analysis of the unsprung motor controller. The finite element model of the random vibration analysis of the unsprung motor controller is as follows: Figure 3 shown.

[0091] Tooling refers to tools, fixtures, molds and other equipment designed and manufactured to achieve specific production processes or operations. In this embodiment, the vibration tooling finite element model is used to simulate the entire vehicle installation structure. It is connected to the motor finite element model and the motor controller finite element model, and provides a vibration environment for the motor finite element model and the motor controller finite element model to simulate the vehicle driving.

[0092] 3) When performing vibration fatigue life assessment, several vibration measurement points are set on the electric drive axle that integrates the motor and motor controller, and corresponding three-axis acceleration sensors are set at each vibration measurement point. The vibration signals of each vibration measurement point are collected by each three-axis acceleration sensor under typical vehicle usage scenarios and stored as corresponding vibration spectra. The specific methods include:

[0093] 3-1) When conducting vibration fatigue life assessment, three-axis acceleration sensors are arranged at vibration measurement points such as the connection between the motor and the final reducer, the left and right leaf spring seats, and the motor connection bracket of a three-in-one electric drive axle of a certain type of micro new energy commercial vehicle. These three-axis acceleration sensors together form a multi-channel data acquisition device;

[0094] 3-2) Using multi-channel data acquisition equipment and a dedicated laptop computer, build a synchronous data acquisition system for vibration spectrum data;

[0095] 3-3) In typical vehicle usage scenarios, such as comprehensive automotive test sites or user scenarios (urban roads, highways, rural roads, and town roads), use the vibration spectrum data synchronization data acquisition system (that is, using each three-axis acceleration sensor) to collect vibration signals at each vibration measurement point in typical vehicle usage scenarios;

[0096] 3-4) The collected vibration signal is subjected to operations such as removing singular values, filtering, and resampling. After completion, the signal is tested for ergodicity, stationarity, and normality, and stored as the corresponding vibration spectrum.

[0097] 4) Using all vibration spectra, the specific methods for establishing a normalized spectrum include:

[0098] 4-1) Calculating the power spectrum density of each vibration spectrum and merging each power spectrum density into a feature sample includes:

[0099] 4-1-1) Calculate the power spectrum density of each vibration spectrum respectively, and record the power spectrum density PSD as Gk(i,j) (i=1,2,…,L; j=1,2,…,M; k=1,2,…,N), where L is the number of vibration signal measurement points, M is the sample size, and N is the number of spectral lines; calculate the root mean square RMS(i,j) of each power spectrum density, as well as the mean and variance of each root mean square;

[0100] The mean and variance of each RMS are calculated according to the following formula:

[0101]

[0102] Where RMS(i,j) is the root mean square of the power spectrum density, is the mean of the root mean square of the power spectral density, is the variance of the root mean square of the power spectral density, M is the sample size;

[0103] 4-1-2) Based on the means and variances calculated in step 4-1-1), calculate the F statistic and t statistic of the power spectral density corresponding to each mean and variance;

[0104] Calculate two statistics F(i,m) and t(i,m) according to the following formula.

[0105]

[0106] Where i is the i-th vibration measurement point, m is the m-th vibration measurement point (i = 1, 2, ..., L; m = 1, 2, ..., L; i ≠ m); F(i, m) is the F statistic determined by the measurement point i and the measurement point m; t(i, m) is the t statistic determined by the measurement point i and the measurement point m; is the variance of the root mean square of the power spectrum density corresponding to the measurement point i; is the variance of the root mean square of the power spectrum density corresponding to the measurement point m; is the mean of the root mean square of the power spectrum density corresponding to the measurement point i; is the mean of the root mean square of the power spectrum density corresponding to the measurement point m; M is the sample size;

[0107] 4-1-3) Perform F-test and t-test on the F-statistic and t-statistic calculated in step 4-1-2), and filter each power spectrum density in the following way to obtain several power spectrum densities that meet the merging conditions:

[0108] If the F statistic passes the F test and the t statistic passes the t test, the corresponding power spectrum density is retained;

[0109] If the F statistic fails the F test, or the t statistic fails the t test, the corresponding power spectrum density is discarded;

[0110] The specific formulas of the F test and t test are:

[0111]

[0112] Where, F (M-1,M-1)；α / 2 is the α / 2 quantile of the F distribution with degrees of freedom (M-1,M-1), F(i,m) is the F statistic, F (M-1,M-1)；(1-α / 2) is the (1-α / 2) quantile of the F distribution with degrees of freedom (M-1,M-1), t(i,m) is the t statistic, t 2(M-1)；(1-α / 2) is a t-distribution (1-α / 2) with 2(M-1) degrees of freedom.

[0113] 4-1-4) Merge the power spectrum density obtained in step 4-1-3) to obtain the feature sample G(p,q) (p=1,2,…,P;q=1,2,…,Q p ), P is the number of feature samples, Q p is the characteristic sample size.

[0114] 4-2) Set the fitting curve and use the fitting curve to normalize the characteristic samples according to the following formula to obtain the normalized characteristic samples:

[0115] w=γ+ηk i (x,λ,ε) (4)

[0116] In the formula, w is the normalized sample, x is the characteristic sample, k i (x,λ,ε) is the fitting curve, η is the first curve parameter, λ is the second curve parameter, and ε is the third curve parameter;

[0117] The fitting curve can be calculated by the following formula:

[0118]

[0119] Where 3z is the first standard normal distribution quantile, z is the second standard normal distribution quantile, -3z is the third standard normal distribution quantile, -z is the fourth standard normal distribution quantile (-3z, -z, z, 3z are four equally spaced standard normal distribution quantiles. The quantiles of the characteristic samples are obtained by these four quantiles, and the characteristic sample quantiles are obtained by combining the interpolation method). 3z is the first characteristic sample quantile, X z is the second characteristic sample quantile, X -3z is the third characteristic sample quantile, X -z is the fourth characteristic sample quantile.

[0120] In this embodiment, the value range of z is 0.25 to 1.25, and the optimal z value is obtained through grid search.

[0121] The γ, η, λ, and ε in formula (5) are calculated as follows:

[0122] When mn / p 2 When <1:

[0123]

[0124] When mn / p 2 When >1:

[0125]

[0126] When mn / p 2 =1:

[0127]

[0128] 4-3) Calculate the mean, variance, and upper tolerance coefficient of the normalized feature sample, and calculate the upper limit of the normalized feature sample based on the mean, variance, and upper tolerance coefficient. The specific methods include:

[0129] 4-3-1) Estimate the mean and variance of the normalized feature sample w according to the following formula:

[0130]

[0131] Where, is the normalized sample mean of the kth spectral line of the pth characteristic sample, is the normalized sample variance of the kth spectral line of the pth characteristic sample, Q p is the feature sample size, p is the number of feature samples, w k (p,q) is the sample after the characteristic sample is normalized;

[0132] 4-3-2) Determine the upper tolerance limit coefficient F according to the following formula:

[0133]

[0134] Where, t Qp-1；(1-α) The degrees of freedom are Q p The 1-α quantile of the t-distribution with a value of -1, χ 2 Qp-1；α is the α quantile of the chi-square distribution with 1 degree of freedom; Z β is the quantile with cumulative probability β on the standard normal distribution, Q p is the characteristic sample size.

[0135] 4-3-3) Estimate the upper limit of the normalized characteristic sample according to the following formula:

[0136]

[0137] Where, is the upper limit of the pth spectral line of the pth normalized characteristic sample, is the normalized sample mean of the kth spectrum line of the pth characteristic sample, F is the upper limit coefficient of the tolerance, S k (p) is the normalized sample standard deviation of the kth spectral line of the pth characteristic sample;

[0138] 4-4) Substitute the upper limit of the normalized characteristic sample obtained in step 4-3) into the following equations to obtain the measured spectrum with frequency distribution characteristics in the three directions of X, Y, and Z:

[0139] mn / p2 <1 case:

[0140]

[0141] For mn / p 2 >1, the inverse transformation is as follows:

[0142]

[0143] For mn / p 2 =1, the inverse transformation is as follows:

[0144]

[0145] 4-5) Based on the frequency domain distribution characteristics of the measured spectrum, the measured spectrum is divided into frequency bands, and the measured spectrum after frequency band division is converted into a standard spectrum, and the loading time of the standard spectrum is compressed. The specific methods include:

[0146] 4-5-1) Divide the frequency bands according to the frequency domain distribution characteristics of the measured spectrum, and then convert the measured spectrum into a standard spectrum according to the following formula:

[0147]

[0148] Where U is the normalized spectrum, σ is the RMS value of the measured spectrum in a certain frequency band, and Δf is the bandwidth of a certain frequency band.

[0149] 4-5-2) Compress the loading time of the standard spectrum in the following way:

[0150]

[0151] Where W1 is the normalized spectrum after compression, W2 is the normalized spectrum before compression, T1 is the operating time after compression, T2 is the road test time at the vehicle proving ground or the user's full life usage time, and m is the slope of the material SN curve.

[0152] 5) Input the canonical spectrum as the simulated load into the finite element model for random vibration analysis of the unsprung motor controller, perform random vibration intensity simulation on the finite element model for random vibration analysis of the unsprung motor controller, and optimize the motor controller structure based on the simulation results until the strength of the motor controller meets the requirements. Specific methods include:

[0153] 5-1) Use tensile testing to obtain the tensile strength limit of the motor controller material;

[0154] 5-2) Using the canonical spectrum as the simulation load, input it into the finite element model for random vibration analysis of the unsprung motor controller, perform random vibration intensity simulation on the finite element model for random vibration analysis of the unsprung motor controller, and obtain the root mean square stress value;

[0155] 5-3) Based on the RMS stress value and the material tensile strength, use the following formula to determine whether the motor controller's vibration intensity is qualified:

[0156] If 3×root mean square stress value is less than the tensile limit of the material, the vibration strength of the motor controller is qualified;

[0157] If the 3×root mean square stress value is ≥ the material tensile limit, the vibration strength of the motor controller is unqualified and the structural design of the motor controller is optimized.

[0158] 6) Performing a swept frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain a stress-frequency response relationship. The obtained stress-frequency response relationship is imported into fatigue analysis software. The standard spectrum is input into the fatigue analysis software as a load. The fatigue life of the unsprung motor controller is then calculated based on the stress-frequency response relationship. The specific method includes:

[0159] 6-1) Setting the amplitude and frequency range of the sweep vibration simulation, wherein the amplitude is 1g and the frequency is in the range of 0 to 2000 Hz;

[0160] 6-2) Perform a swept frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain the stress-frequency response results of the unsprung motor controller;

[0161] 6-3) Importing the swept frequency vibration stress-frequency response results obtained in step 5-2) into fatigue analysis software, and using the canonical spectrum as the load input and the canonical spectrum time as the loading time;

[0162] 6-4) Estimate the material's stress-life curve (SN curve) in fatigue analysis software based on the material's tensile strength. Calculate the fatigue life of the unsprung motor controller using the Dirlik model based on the stress-life curve.

[0163] This embodiment also involves a random vibration fatigue bench test. The random vibration test site is as follows: Figure 4 As shown in the figure, the random vibration fatigue bench test is used to further verify the random vibration fatigue life analysis results of the unsprung motor controller, thereby improving the accuracy and effectiveness of the simulation results.

[0164] The random vibration fatigue bench test specifically includes:

[0165] S1) Prepare the test sample for the random vibration bench test, i.e., the motor-motor controller two-in-one system, and inspect the system in terms of appearance, function, and performance to ensure that all indicators of the test sample meet the requirements;

[0166] S2) The normalized spectrum is used as a load and input into the electromagnetic vibration table (actually, it is input into the host computer of the electromagnetic vibration table), and the time of the normalized spectrum is used as the loading time of the electromagnetic vibration table;

[0167] S3) performing a random vibration fatigue test on the motor-motor controller two-in-one system using an electromagnetic vibration test bench;

[0168] S4) Check the appearance, function, and performance indicators of the motor-motor controller two-in-one system. If the inspection fails, the structural optimization design of the motor controller is still required.

[0169] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications made to the present invention by those skilled in the art without departing from the spirit of the present invention shall fall within the scope of protection of the present invention.

Claims

1. A vibration fatigue life assessment method for an unsprung motor controller, characterized in that: The following steps are involved: 1) Design and manufacture vibration tooling for vibration fatigue life assessment based on the motor controller and the corresponding motor; 2) Before vibration fatigue life assessment, corresponding finite element models are established for the vibration fixture, motor, and motor controller. These finite element models are then combined into a finite element model for random vibration analysis of the unsprung motor controller. 3) When performing vibration fatigue life assessment, several vibration measurement points are set on the electric drive axle, and corresponding three-axis acceleration sensors are set at each vibration measurement point to obtain several vibration spectra; 4) Using all vibration spectra, establish a standard spectrum; 5) Input the canonical spectrum as the simulated load into the finite element model for random vibration analysis of the unsprung motor controller to perform random vibration intensity simulation. Based on the simulation results, the motor controller is structurally optimized until the strength of the motor controller meets the requirements. 6) Perform a swept-frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain the stress-frequency response relationship. This stress-frequency response relationship is then imported into fatigue analysis software. The canonical spectrum is then input as the load into the fatigue analysis software to calculate the fatigue life of the unsprung motor controller.

2. The vibration fatigue life assessment method according to claim 1, characterized in that: In step 4), the specific method of establishing a normalized spectrum using all vibration spectra includes: 4-1) Calculate the power spectrum density of each vibration spectrum and merge each power spectrum density into a feature sample; 4-2) Set a fitting curve and use the fitting curve to normalize the characteristic samples according to the following formula to obtain normalized characteristic samples; 4-3) Calculating the mean, variance, and upper tolerance coefficient of the normalized characteristic sample, and calculating the upper limit of the normalized characteristic sample based on the mean, variance, and upper tolerance coefficient; 4-4) Calculate the measured spectrum based on the upper limit of the normalized characteristic sample; 4-5) According to the frequency domain distribution characteristics of the measured spectrum, the measured spectrum is divided into frequency bands, and the measured spectrum after frequency band division is converted into a standard spectrum, and the loading time of the standard spectrum is compressed.

3. The vibration fatigue life assessment method according to claim 1, characterized in that: In step 4-1), the specific method of calculating the power spectrum density of each vibration spectrum and merging each power spectrum density into a feature sample includes: 4-1-1) Calculate the power spectrum density of each vibration spectrum, and calculate the root mean square of each power spectrum density, as well as the mean and variance of each root mean square; 4-1-2) Based on the means and variances calculated in step 4-1-1), calculate the F statistic and t statistic of the power spectral density corresponding to each mean and variance; 4-1-3) Perform F-test and t-test on the F-statistic and t-statistic calculated in step 4-1-2), and filter each power spectrum density in the following way to obtain several power spectrum densities that meet the merging conditions: If the F statistic passes the F test and the t statistic passes the t test, the corresponding power spectrum density is retained; If the F statistic fails the F test, or the t statistic fails the t test, the corresponding power spectrum density is discarded; 4-1-4) The power spectrum density obtained in step 4-1-3) is merged to obtain feature samples.

4. The vibration fatigue life assessment method according to claim 3, characterized in that: The specific formulas of the F test and t test are: Where, F (M-1,M-1)；α / 2 is the α / 2 quantile of the F distribution with degrees of freedom (M-1,M-1), F(i,m) is the F statistic, F (M-1,M-1)；(1-α / 2) is the (1-α / 2) quantile of the F distribution with (M-1,M-1) degrees of freedom, t(i,m) is the t statistic, t 2(M-1)；(1-α / 2) is a t-distribution (1-α / 2) with 2(M-1) degrees of freedom.

5. The vibration fatigue life assessment method according to claim 1, characterized in that: The specific method of performing random vibration intensity simulation on the finite element model of random vibration analysis of the unsprung motor controller includes: 5-1) Use tensile testing to obtain the tensile strength limit of the motor controller material; 5-2) Using the canonical spectrum as the simulation load, input it into the finite element model for random vibration analysis of the unsprung motor controller, perform random vibration intensity simulation on the finite element model for random vibration analysis of the unsprung motor controller, and obtain the root mean square stress value; 5-3) Based on the RMS stress value and the material tensile strength, use the following formula to determine whether the motor controller's vibration intensity is qualified: If 3×root mean square stress value is less than the tensile limit of the material, the vibration strength of the motor controller is qualified; If the 3×root mean square stress value is ≥ the material tensile limit, the vibration strength of the motor controller is unqualified and the structural design of the motor controller is optimized.

6. The vibration fatigue life assessment method according to claim 1, characterized in that: The specific method of performing random vibration fatigue life assessment on the random vibration analysis finite element model of the unsprung motor controller includes: 6-1) Set the amplitude and frequency range of the swept frequency vibration simulation; 6-2) Perform a swept frequency vibration simulation on the finite element model for random vibration analysis of the unsprung motor controller to obtain the stress-frequency response results of the unsprung motor controller; 6-3) Importing the swept frequency vibration stress-frequency response results obtained in step 5-2) into fatigue analysis software, and using the canonical spectrum as the load input and the canonical spectrum time as the loading time; 6-4) Estimate the material's stress-life curve using fatigue analysis software based on the material's tensile strength. Calculate the fatigue life of the unsprung motor controller based on the stress-life curve.

7. The vibration fatigue life assessment method according to claim 1, characterized in that: It also involves a random vibration fatigue bench test to further verify the results of the random vibration fatigue life analysis of the unsprung motor controller.

8. The vibration fatigue life assessment method according to claim 1, characterized in that: The several vibration measuring points include a vibration measuring point at the connection between the motor and the main reducer, a vibration measuring point at the left and right leaf spring seats, and a vibration measuring point at the motor connection bracket.

Images