CN116011112A - Aeroengine bearing fault data simulation method based on transmission characteristics of casing - Google Patents
Aeroengine bearing fault data simulation method based on transmission characteristics of casing Download PDFInfo
- Publication number
- CN116011112A CN116011112A CN202310087958.8A CN202310087958A CN116011112A CN 116011112 A CN116011112 A CN 116011112A CN 202310087958 A CN202310087958 A CN 202310087958A CN 116011112 A CN116011112 A CN 116011112A
- Authority
- CN
- China
- Prior art keywords
- bearing
- fault
- casing
- excitation
- simulation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses an aeroengine bearing fault data simulation method based on a casing transmission characteristic, which comprises the following steps: simulating and extracting the contact force between the bearing outer ring and the bearing seat under different bearing fault modes based on the dynamic characteristics of the bearing; simplifying contact force based on different failure mode mechanisms of the bearing, extracting periodic impact excitation force and simplifying the periodic impact excitation force; applying bearing fault equivalent simplified single excitation force, and extracting a casing measuring point vibration response based on a casing transfer characteristic mapping mechanism; recombining and superposing vibration responses of the measuring points in a time domain based on the working principle of the rolling bearing; based on a fault mechanism and simulation time length, performing response matrix reconstruction and signal modulation, and fusing actual measurement background noise of an engine to form a fault case; and carrying out envelope spectrum analysis on the numerical simulation fault case, extracting peak characteristic frequency in the envelope spectrum, and comparing the error between the peak frequency in the envelope spectrum and the theoretical characteristic frequency to obtain the accuracy of the fault simulation method.
Description
Technical Field
The invention relates to the technical field of aero-engine bearing fault simulation, in particular to an aero-engine bearing fault data simulation method based on a casing transmission characteristic.
Background
The bearing is used as a supporting component of an aeroengine rotor system, and the dynamic characteristic of the bearing has a remarkable influence on the working performance of the rotor, and even directly influences the functions of the whole aeroengine. The bearing works under the conditions of high temperature, high speed, high DN value and complex load change for a long time, so that the fault is frequent. The related data show that the failure mainly comprises rotor component and bearing faults, and accounts for more than 80% of the serious mechanical failures of the engine. In recent years, a deep learning method based on big data provides an important way for intelligent fault diagnosis of mechanical equipment. However, in engineering practice, the monitoring data of the mechanical equipment have low availability and low value density, which is due to the fact that the fault characteristics of the bearings are weak, the sensors are difficult to install, the signal acquisition difficulty is high, and the like, so that insufficient information can be generated in the fault data collected and marked for each bearing state. Specifically, when the faulty sample is far less than the normal sample, a data distribution imbalance phenomenon occurs; and when the absolute number of the various bearing data is small, a small sample is presented. Severely hampers engineering applications of engine bearing fault diagnosis techniques.
When bearing faults occur early, the fault characteristics are weak and are accompanied with strong noise background, and vibration signals are often comprehensive expression of vibration responses of a plurality of parts, so that the uncertainty of transmission paths of the sensors and attenuation and distortion of signal energy are more complex in bearing weak fault characteristics in a casing vibration signal, and the vibration signals containing bearing fault characteristic information are difficult to distinguish from healthy vibration signals, so that the difficulty of bearing fault diagnosis is increased. The deep learning method can be used for adaptively extracting representative fault characteristics from a large amount of data by establishing a deep structure model, so that the problems of unbalanced engine bearing fault samples and small samples are needed to be solved.
At present, students at home and abroad mainly study the problem of sample imbalance from two aspects of data and algorithm. The former aims to expand the number of minority class samples by means of sample resampling or data generation, while the latter aims to increase the sensitivity and punishment of the model to minority classes to reduce diagnostic errors or optimize the classifier. Although the method has a certain effect, the resampling method also causes overgeneralization, overgeneration or change of the distribution condition of the original data when a new sample is synthesized, and the problems of fuzzy generated sample and lack of texture details exist. The algorithm is often difficult to reasonably set specific cost parameters, and when the data imbalance degree is too high, the lifting effect of the model is limited. Aiming at the small sample classification problem in each field, the method mainly comprises four categories of data enhancement, migration learning, meta learning and metric learning. The above method can improve the fault diagnosis performance under a small sample, but has a plurality of problems. For example, data enhancement methods may introduce noisy data; the transfer learning method essentially still needs a large amount of auxiliary data as a premise; the complexity of the meta learning method is high, and the related technology is still immature; and metric learning is less accurate in cases where there are few training samples.
Calculating the transfer function of a linear system is a common method of describing the vibration transfer characteristics of a structure. For a linear time-invariant system, its system output response is equal to the convolution of the system input and the system impulse response. According to the time domain convolution theorem, the spectral density of the output signal of the system is known to be equal to the product of the frequency spectrum of the input signal and the impulse response frequency spectrum. The fourier transform of the system impulse response is the frequency domain transfer function of the system. When people study the transfer characteristic, the transfer characteristic is often more focused on the frequency domain transfer characteristic, and a series of calculation methods for calculating the frequency response function of the linear system are generated, and mainly comprise an amplitude method, a cross-spectrum method, an inversion method, a least square method, a scale factor method, an arithmetic average method and the like. However, for aeroengine bearing faults, the excitation and response of the aeroengine bearing faults have obvious time domain periodic characteristics, so that the simulation of the bearing faults only in the frequency domain is difficult to ensure that the bearing fault mechanism is met. At present, the research on how to map the relation between excitation and response by utilizing the vibration transmission characteristics of the aero-engine casing in the time domain so as to accurately and efficiently simulate the bearing faults still has certain defects.
Disclosure of Invention
In order to solve the problems in the prior art, the embodiment of the invention provides an aeroengine bearing fault data simulation method based on a casing transmission characteristic. The technical scheme is as follows:
in one aspect, an aeroengine bearing fault data simulation method based on a casing transmission characteristic is provided, including:
step 1): simulating and extracting the contact force between the bearing outer ring and the bearing seat under different bearing fault modes based on the dynamic characteristics of the bearing;
step 2): simplifying contact force based on different failure mode mechanisms of the bearing, extracting periodic impact excitation force and simplifying the periodic impact excitation force;
step 3): applying bearing fault equivalent simplified single excitation force, and extracting a casing measuring point vibration response based on a casing transfer characteristic mapping mechanism;
step 4): recombining and superposing vibration responses of the measuring points in a time domain based on the working principle of the rolling bearing;
step 5): based on a fault mechanism and simulation time length, performing response matrix reconstruction and signal modulation, and fusing actual measurement background noise of an engine to form a fault case;
step 6): and carrying out envelope spectrum analysis on the numerical simulation fault case, extracting peak characteristic frequency in the envelope spectrum, and comparing the error between the peak frequency in the envelope spectrum and the theoretical characteristic frequency to obtain the accuracy of the fault simulation method.
Further, in step 1), the method includes the steps of simulating and extracting the contact force between the bearing outer ring and the bearing seat in different bearing failure modes based on the dynamic characteristics of the bearing, specifically:
setting faults of different components of the bearing by using a finite element model or a multi-body dynamics model and the like, and referring to actual added constraint and boundary conditions of engineering so as to finish dynamic characteristic simulation of the bearing faults under different modes; and extracting the contact force between the bearing outer ring and the bearing seat under different bearing fault modes through post-treatment.
Further, in step 2), the contact force is simplified based on different failure mode mechanisms of the bearing, and the periodic impact excitation force is extracted and simplified, specifically:
calculating bearing fault excitation force periods according to dynamic characteristic simulation data characteristic frequencies of different bearing fault modes, judging the moment when first contact force is generated between balls and defects, extracting contact force in the corresponding periods to obtain fault excitation force of different bearing fault modes, and simplifying the bearing fault excitation force into a value F which is equal to the magnitude average value of the fault excitation force 1 The excitation matrix M is of constant amplitude with periodic time domain impulse characteristics.
Specifically, the time when the first contact force is generated between the ball and the defect is determined specifically as follows:
and searching a sampling point with the largest absolute value of the contact force between the bearing and the bearing seat in the first fault excitation force period of the bearing, namely recording as the beginning of the contact force between the ball and the defect.
Further, in step 3), the bearing failure applying equivalent simplifies the single excitation force, and extracts the vibration response of the receiver measuring point based on the receiver transfer characteristic mapping mechanism, specifically:
based on the simplified constant amplitude periodic time domain excitation matrix M, single pulse excitation N is extracted and applied to a three-dimensional entity finite element model of an aeroengine casing bearing or a high-fidelity casing to carry out transient dynamics analysis; and mapping the relation between bearing fault excitation and measuring point vibration response based on the transmission characteristic of the engine casing, and extracting three measuring point vibration acceleration data of the fan casing, the intermediate casing and the rear casing of the turbine by combining the actual monitoring position of the engine as the basis of bearing fault numerical simulation.
Specifically, the response of the case measuring point under single pulse excitation is equivalent to Shan Cizhou failure excitation based on visual representation of the case transfer characteristic, and meanwhile, the engine case transfer characteristic is calculated through a simulation model or is obtained by testing by utilizing an actual case structure.
Further, in step 4), the vibration response of the measuring point is recombined and overlapped in the time domain based on the working principle of the rolling bearing, specifically:
based on the bearing fault excitation characteristics, grouping the case measuring point signals according to sampling points contained in a single excitation period of bearing faults in different fault modes to obtain a measuring point response matrix P under single pulse excitation; and then, according to the periodic working principle of the rolling bearing, carrying out signal superposition reconstruction calculation based on the measuring point response matrix P under single pulse excitation to obtain a vibration response matrix Q of the measuring point of the time box when the bearing rotates for one circle.
Further, in step 5), based on the failure mechanism and the simulation time length, the response matrix reconstruction and the signal modulation are performed, and the actual measurement background noise of the engine is fused to form a failure case, which specifically includes:
calculating the bearing working revolution n by using the simulation case duration/the bearing working 1-week duration, copying, expanding and reconstructing a response matrix Q, modulating the reconstructed matrix according to a bearing fault mechanism, and finally fusing the actually measured background noise of the engine to form a bearing fault case under a corresponding fault mode.
Specifically, reconstructing and expanding the extracted vibration acceleration signals of the case measuring points, and reconstructing a bearing fault response matrix based on different bearing fault mode fault mechanisms, exciting force periods and sampling points in the periods to obtain reconstructed vibration acceleration time domain waveforms under different bearing fault modes.
Further, in step 6), performing envelope spectrum analysis on the digital simulation fault case, extracting peak characteristic frequency in the envelope spectrum, and comparing the error between the peak frequency in the envelope spectrum and the theoretical characteristic frequency to obtain the accuracy of the fault simulation method, specifically:
analyzing the bearing fault numerical simulation case by utilizing the envelope spectrum, and calculating the corresponding fault characteristic frequency according to the bearing fault characteristic frequency calculation formula, wherein the method comprises the following steps:
(1) Bearing outer ring failure:
(2) Bearing inner ring failure:
(3) Bearing rolling element failure:
(4) Bearing cage failure:
and finally, calculating the error between the peak frequency in the envelope spectrum and the theoretical characteristic frequency to obtain the accuracy of the fault simulation method.
The technical scheme provided by the embodiment of the invention has the beneficial effects that:
the invention provides an aeroengine bearing fault data simulation method based on a casing transfer characteristic, which is characterized in that dynamic characteristic simulation of different bearing fault modes is carried out based on a finite element model or a multi-body dynamic model and the like, contact force between a bearing outer ring and a bearing seat is extracted, fault excitation force is equivalently simplified based on a bearing fault mechanism, the fault excitation force is applied to a three-dimensional entity finite element model or an actual engine casing of an aeroengine high-fidelity casing, a measuring point vibration response is extracted based on a casing transfer characteristic mapping mechanism, a rolling bearing working principle is adopted, a response matrix is reconstructed and modulated by combining the fault mechanism, and finally actual measurement background noise is fused.
The invention completes the work of extracting and simplifying the excitation force of different bearing fault modes, mapping the transmission characteristic of the case, reconstructing the matrix, modulating the signal and the like, and successfully generates the engine bearing fault numerical simulation case based on the bearing fault mechanism and the transmission characteristic of the engine case. The method can be used for learning bearing fault characteristics by a computer, and further provides support for fault characteristic transmission, mechanism analysis and fault diagnosis.
The invention provides an aeroengine bearing fault data simulation method based on a casing transmission characteristic, which can extract bearing fault excitation force characteristics by combining a bearing fault mechanism, extract measuring point vibration response by depending on the aeroengine casing vibration transmission characteristic, reconstruct a response matrix based on a rolling bearing working principle and the fault mechanism in a time domain, and integrate actual measurement background noise to form a fault case. The method can ensure the effectiveness and accuracy of each fault mode of the bearing, can meet the short-time generation requirement of a large number of fault cases, and can effectively solve the problems of unbalanced fault samples and small samples of the engine bearing, thereby better promoting the rapid development of the intelligent diagnosis system of the aeroengine.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of an aero-engine bearing fault data simulation method based on a time domain impact mechanism and a casing vibration transmission characteristic;
FIG. 2 is a diagram of a contact between an outer ring of a bearing and a bearing seat based on a multi-body dynamics model simulation bearing inner ring fault extraction in an embodiment of the invention;
FIG. 3 is a simplified bearing inner race fault excitation diagram based on a fault mechanism in an embodiment of the present invention;
FIG. 4 is a schematic representation of a single bearing failure excitation based on an aircraft engine case transfer characteristic mapping mechanism in an embodiment of the present invention;
FIG. 5 is a graph showing the vibration acceleration response of the intermediate case measurement point for one week of bearing operation after reconstruction in an embodiment of the present invention;
FIG. 6 is a time domain diagram of the measured background noise of an engine according to an embodiment of the present invention;
FIG. 7 is a time domain waveform diagram of a reconstructed bearing inner ring fault numerical simulation case in an embodiment of the present invention;
FIG. 8 is a graph of a result of analysis of an envelope spectrum of a bearing inner race fault numerical simulation case in an embodiment of the present invention;
fig. 9 is a schematic structural diagram of an aero-engine and a schematic diagram of a transmission characteristic of a casing according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the embodiments of the present invention will be described in further detail with reference to the accompanying drawings.
The invention provides an aeroengine bearing fault data simulation method based on a casing transmission characteristic, which comprises the following steps:
s1: simulating and extracting the contact force between the bearing outer ring and the bearing seat under different bearing fault modes based on the dynamic characteristics of the bearing;
s2: simplifying contact force based on different failure mode mechanisms of the bearing, extracting periodic impact excitation force and simplifying the periodic impact excitation force;
s3: applying bearing fault equivalent simplified single excitation force, and extracting a casing measuring point vibration response based on a casing transfer characteristic mapping mechanism;
s4: recombining and superposing vibration responses of the measuring points in a time domain based on the working principle of the rolling bearing;
s5: based on a fault mechanism and simulation time length, performing response matrix reconstruction and signal modulation, and fusing actual measurement background noise of an engine to form a fault case;
s6: and carrying out envelope spectrum analysis on the numerical simulation fault case, extracting peak characteristic frequency in the envelope spectrum, and comparing the error between the peak frequency in the envelope spectrum and the theoretical characteristic frequency to obtain the accuracy of the fault simulation method.
Specifically, the aero-engine bearing fault data simulation method is specifically an aero-engine bearing fault data simulation method based on a time domain impact mechanism and a casing vibration transmission characteristic.
Meanwhile, the invention utilizes finite element models or multi-body dynamics models and the like to simulate the bearing fault dynamics based on a bearing fault mechanism, extracts the contact force between the bearing outer ring and the bearing seat, simplifies the treatment, is applied to the actual engine casing bearing or the high-fidelity casing three-dimensional entity finite element model, extracts the casing measuring point vibration response based on a casing transfer characteristic mapping mechanism, then carries out matrix reconstruction and modulation treatment based on a rolling bearing working principle, combines with the bearing fault mechanism, and finally fuses with actual engine background noise, thereby obtaining a bearing fault numerical simulation case based on the fault mechanism and the aeroengine casing transfer characteristic, and is used for learning an aeroengine fault diagnosis expert system.
Examples:
the data of the embodiment are obtained by simulating the fault dynamics characteristics of the inner ring of the bearing based on multi-body contact transient dynamics and simulating the complex transmission characteristics of the casing based on a three-dimensional entity finite element model of the high-fidelity casing of the aeroengine. The bearing is built according to an aeroengine rolling bearing, a rigid bearing and a flexible bearing seat are adopted for coupling mode modeling, groove defects are arranged on the inner ring of the bearing, and the bearing is in a stable working condition in the fault simulation process; the complete machine casing comprises 6 parts of a low-pressure compressor casing, an intermediate casing, a high-pressure compressor casing, a combustion chamber casing, a turbine casing, an afterburner casing and a tail nozzle casing.
A method for simulating aeroengine bearing fault data based on a time domain impact mechanism and a casing vibration transmission characteristic comprises the following specific steps as shown in a figure 1:
s1, setting a bearing inner ring fault by utilizing a multi-body dynamics model, and referring to engineering actual addition constraint and boundary conditions, so as to complete the simulation of the bearing inner ring fault dynamics characteristic. And extracting the contact force between the bearing outer ring and the bearing seat under the fault mode of the bearing inner ring through post-treatment.
In the embodiment, the fault dynamics characteristic of the inner ring of the bearing is simulated by utilizing a multi-body dynamics model, the defect width of the groove of the inner ring of the bearing is set to be 4mm, the depth of the groove of the inner ring of the bearing is set to be 2mm, the driving rotating speed of the inner ring is set to be the full working condition rotating speed, the simulation time is 0.2s, the sampling point is 12810, and the contact force between the outer ring of the bearing and the bearing seat is extracted as shown in the attached figure 2.
S2, calculating a bearing fault excitation force period based on the characteristic frequency of the bearing inner ring fault dynamics characteristic simulation data, judging the moment when the first contact force is generated between the ball and the defect, and extracting the simulation contact force between the bearing outer ring and the bearing seat in the corresponding excitation period under the bearing inner ring fault mode, namely the fault excitation force of the bearing inner ring fault. Then simplifying the fault excitation force of the bearing inner ring into a value F which is the average value of the fault excitation force 1 The excitation matrix M is of constant amplitude with periodic time domain impulse characteristics.
In the embodiment, the characteristic frequency of the multi-body dynamics fault simulation data is 2135Hz, and the corresponding excitation force period is 4.684 multiplied by 10 -4 s. The single excitation force period comprises 30 sampling points, and the 17 th sampling point is determined to be the maximum value in the period after judgment, namely the moment of generating the first contact force is 2.654 multiplied by 10 < -4 > s. The average value of the fault excitation force of the bearing inner ring is 18499N, and the fault excitation force of the bearing inner ring after the simplification treatment is shown in the figure 3.
S3, extracting single pulse excitation N by using a simplified constant-amplitude periodic time domain impulse excitation matrix M, applying the single pulse excitation N to a three-dimensional entity finite element model of the high-fidelity casing of the aeroengine, setting simulation time length and sub-step for transient dynamics analysis. And extracting vibration acceleration data of three measuring points of the fan casing, the intermediate casing and the rear casing of the turbine by combining with an actual monitoring position by means of a casing vibration transmission characteristic mapping mechanism, and taking the vibration acceleration data as a basis of bearing inner ring fault numerical simulation.
In the embodiment of the invention, the simulation time length is set to be 0.1s, and the step length is set to be 1/100 of the excitation period, namely 4.684 multiplied by 10 -5 s, calculating the representation of single bearing inner ring fault equivalent excitation based on the transfer characteristic of the high-fidelity casing three-dimensional entity finite element model by utilizing transient dynamics analysis, and extracting the vibration acceleration response of the intermediate casing measuring point as shown in figure 4.
S4, based on the characteristics of bearing fault excitation, grouping the measuring point signals of the casing by the sampling points contained in a single excitation period of bearing inner ring faults to obtain a measuring point response matrix P under single pulse excitation. And according to the periodic working principle of the rolling bearing, performing signal superposition calculation based on a measuring point response matrix P under single pulse excitation to obtain a response matrix Q of measuring points of the fan casing, the intermediate casing and the rear casing of the turbine when the bearing rotates for one circle.
In the embodiment of the invention, 30 sampling points are included in a single excitation period of the bearing, and the responses of the case measuring points in different excitation periods are overlapped to obtain a time domain waveform diagram of vibration acceleration of the case measuring point of the intermediary in which the bearing works for one circle after reconstruction, and the time domain waveform diagram is shown in figure 5.
S5, calculating the bearing working revolution n by using the simulation case duration/the bearing working revolution duration, copying, expanding and reconstructing the response matrix Q according to the bearing working revolution, modulating the reconstructed matrix and the rotating speed signal by combining a bearing inner ring fault mechanism, and fusing the actually measured background noise of the engine to form a bearing inner ring fault case.
In the embodiment of the invention, fault simulation time length is set to be 0.1s, the working revolution number of the bearing is calculated to be 214 revolutions, then the response matrix of the bearing rotating for one circle is utilized to carry out copying, expanding and reconstructing, and after the frequency conversion component in the normal actual measurement signal of the engine is removed by utilizing a filter, the signal is used as the background noise of the aeroengine as shown in figure 6. And overlapping the bearing inner ring failure simulation case with the modulated reconstruction matrix to obtain the bearing inner ring failure numerical simulation case based on the bearing inner ring failure mechanism and the transmission characteristic of the casing, wherein the time domain waveform is shown in figure 7.
S6, carrying out envelope spectrum analysis on the bearing inner ring fault numerical simulation case, and extracting peak characteristic frequency and theoretical fault characteristic frequency comparison in an envelope spectrum, wherein the method specifically comprises the following steps: and analyzing a bearing inner ring fault numerical simulation case by using the envelope spectrum, calculating a corresponding fault characteristic frequency according to a bearing fault characteristic frequency formula, calculating an error between a peak value frequency in the envelope spectrum and a theoretical characteristic frequency, and obtaining the accuracy of the fault simulation method by representing the error in percentage.
The analysis result of the response envelope spectrum of the case measuring point of the intermediate case of the bearing inner ring fault case is shown in figure 8, and can be seen that 2135Hz and frequency multiplication components thereof exist besides 219.5Hz of the bearing inner ring rotation frequency, and obvious side frequency band components exist on two sides of the bearing inner ring rotation frequency. At the moment, the theoretical value of the rotation frequency of the inner ring of the bearing is 221.7Hz, and the error is 0.99 percent; the theoretical characteristic frequency of the bearing inner ring faults calculated according to the bearing inner ring fault characteristic frequency formula is 2145Hz, and the error is only 0.47%.
The contact force between the bearing inner ring fault bearing outer ring and the bearing seat extracted by the embodiment is shown in fig. 2, and the obvious periodic impact characteristic can be seen from the graph. The simplified exciting force of the bearing inner ring fault processed based on the bearing fault mechanism in the embodiment is shown in fig. 3, and the bearing inner ring fault has the characteristics of periodicity and constant amplitude. Fig. 4 shows a representation diagram of a single bearing fault excitation based on an aircraft engine case transfer characteristic mapping mechanism, namely an intermediate case measurement point vibration acceleration response diagram, which can be used for subsequent reconstruction of bearing fault time domain signals. Fig. 5 is a waveform reconstruction diagram of vibration acceleration time domain of a measuring point of an intermediate case in a single bearing failure excitation bearing working cycle, and can be used for reconstruction of a subsequent bearing inner ring failure case. Fig. 6 is a time domain waveform diagram of the actual measurement signal of the aero-engine, which removes the power frequency and the frequency multiplication component thereof, and can be used as a fault simulation case to be closer to actual background noise. Fig. 7 is a time domain waveform diagram of a bearing inner race fault numerical simulation case, which can be used as a learning sample of an intelligent diagnosis expert system of an aeroengine. FIG. 8 is a result of analysis of a bearing inner ring fault numerical simulation case envelope spectrum that can be used to determine the accuracy of a bearing inner ring fault simulation method. Fig. 9 is a schematic structural diagram of an aeroengine and a schematic diagram of a transmission characteristic of a casing, which can illustrate that relevant information at the bearing position of the engine can be mapped to a measuring point on the surface of the casing through the transmission characteristic of the thin-wall casing, and is a key for carrying out bearing fault numerical simulation based on a time domain impact mechanism.
The embodiment of the invention provides an aeroengine bearing fault data simulation method based on the transmission characteristics of a casing, which can generate a bearing fault numerical simulation case based on a bearing fault mechanism and the transmission characteristics of the engine casing, meets the bearing fault characteristics through envelope spectrum analysis, can serve for deep learning of an intelligent diagnosis system of the aeroengine, provides an effective solution for the problems of unbalance samples and small samples of actual measurement fault cases of the engine in engineering practice, and is favorable for promoting progress of an aeroengine fault diagnosis expert system.
The foregoing is only illustrative of the present invention and is not to be construed as limiting thereof, but rather as various modifications, equivalent arrangements, improvements, etc., within the spirit and principles of the present invention.
Claims (10)
1. The aeroengine bearing fault data simulation method based on the transmission characteristics of the case is characterized by comprising the following steps of:
step 1): simulating and extracting the contact force between the bearing outer ring and the bearing seat under different bearing fault modes based on the dynamic characteristics of the bearing;
step 2): simplifying contact force based on different failure mode mechanisms of the bearing, extracting periodic impact excitation force and simplifying the periodic impact excitation force;
step 3): applying bearing fault equivalent simplified single excitation force, and extracting a casing measuring point vibration response based on a casing transfer characteristic mapping mechanism;
step 4): recombining and superposing vibration responses of the measuring points in a time domain based on the working principle of the rolling bearing;
step 5): based on a fault mechanism and simulation time length, performing response matrix reconstruction and signal modulation, and fusing actual measurement background noise of an engine to form a fault case;
step 6): and carrying out envelope spectrum analysis on the numerical simulation fault case, extracting peak characteristic frequency in the envelope spectrum, and comparing the error between the peak frequency in the envelope spectrum and the theoretical characteristic frequency to obtain the accuracy of the fault simulation method.
2. The aeroengine bearing fault data simulation method based on the transmission characteristics of the casing as claimed in claim 1, wherein in step 1), the method for simulating and extracting the contact force between the bearing outer ring and the bearing seat in different bearing fault modes based on the bearing dynamics characteristics is specifically as follows:
setting faults of different components of the bearing by using a finite element model or a multi-body dynamics model and the like, and referring to actual added constraint and boundary conditions of engineering so as to finish dynamic characteristic simulation of the bearing faults under different modes; and extracting the contact force between the bearing outer ring and the bearing seat under different bearing fault modes through post-treatment.
3. The aeroengine bearing fault data simulation method based on the transmission characteristics of the casing as claimed in claim 2, wherein in the step 2), the contact force is simplified based on different fault mode mechanisms of the bearing, and the periodic impact excitation force is extracted and simplified, specifically:
calculating bearing fault excitation force periods according to dynamic characteristic simulation data characteristic frequencies of different bearing fault modes, judging the moment when first contact force is generated between balls and defects, extracting contact force in the corresponding periods to obtain fault excitation force of different bearing fault modes, and simplifying the bearing fault excitation force into a value F which is equal to the magnitude average value of the fault excitation force 1 The excitation matrix M is of constant amplitude with periodic time domain impulse characteristics.
4. The method for simulating the bearing failure data of the aero-engine based on the transmission characteristic of the casing according to claim 3, wherein the moment when the first contact force is generated between the ball and the defect is determined specifically as follows:
and searching a sampling point with the largest absolute value of the contact force between the bearing and the bearing seat in the first fault excitation force period of the bearing, namely recording as the beginning of the contact force between the ball and the defect.
5. The aeroengine bearing fault data simulation method based on the transmission characteristics of the casing as claimed in claim 4, wherein in the step 3), the bearing fault applying equivalent simplified single excitation force, the casing measuring point vibration response is extracted based on the mapping mechanism of the transmission characteristics of the casing, specifically:
based on the simplified constant amplitude periodic time domain excitation matrix M, single pulse excitation N is extracted and applied to a three-dimensional entity finite element model of an aeroengine casing bearing or a high-fidelity casing to carry out transient dynamics analysis; and mapping the relation between bearing fault excitation and measuring point vibration response based on the transmission characteristic of the engine casing, and extracting three measuring point vibration acceleration data of the fan casing, the intermediate casing and the rear casing of the turbine by combining the actual monitoring position of the engine as the basis of bearing fault numerical simulation.
6. The aeroengine bearing fault data simulation method based on the casing transfer characteristic according to claim 5, wherein the casing measurement point response under single pulse excitation is equivalent to Shan Cizhou bearing fault excitation based on visual representation of the casing transfer characteristic, and the engine casing transfer characteristic is obtained through calculation of a simulation model or by testing by using an actual casing structure.
7. The aeroengine bearing fault data simulation method based on the transmission characteristics of the casing as claimed in claim 6, wherein in the step 4), the rolling bearing working principle is based on the recombination and superposition of the vibration responses of the measuring points in a time domain, specifically:
based on the bearing fault excitation characteristics, grouping the case measuring point signals according to sampling points contained in a single excitation period of bearing faults in different fault modes to obtain a measuring point response matrix P under single pulse excitation; and then, according to the periodic working principle of the rolling bearing, carrying out signal superposition reconstruction calculation based on the measuring point response matrix P under single pulse excitation to obtain a vibration response matrix Q of the measuring point of the time box when the bearing rotates for one circle.
8. The method for simulating aeroengine bearing fault data based on the transmission characteristics of a casing according to claim 7, wherein in step 5), based on a fault mechanism and simulation time length, response matrix reconstruction and signal modulation are performed, and engine actual measurement background noise is fused to form a fault case, which specifically comprises:
calculating the bearing working revolution n by using the simulation case duration/the bearing working 1-week duration, copying, expanding and reconstructing a response matrix Q, modulating the reconstructed matrix according to a bearing fault mechanism, and finally fusing the actually measured background noise of the engine to form a bearing fault case under a corresponding fault mode.
9. The aeroengine bearing fault data simulation method based on the transmission characteristics of the casing according to claim 8, wherein the reconstruction and expansion of the extracted vibration acceleration signals of the casing measuring points are realized based on different bearing fault mode fault mechanisms and periods of exciting forces and sampling points in the periods, so as to obtain the reconstructed vibration acceleration time domain waveforms in different bearing fault modes.
10. The aeroengine bearing fault data simulation method based on the casing transfer characteristic of claim 9, wherein in step 6), envelope spectrum analysis is performed on a digital simulation fault case, peak characteristic frequency in the envelope spectrum is extracted, and errors of the peak frequency and theoretical characteristic frequency in the envelope spectrum are compared to obtain the accuracy of the fault simulation method, and the method is specifically as follows:
analyzing the bearing fault numerical simulation case by utilizing the envelope spectrum, and calculating the corresponding fault characteristic frequency according to the bearing fault characteristic frequency calculation formula, wherein the method comprises the following steps:
(1) Bearing outer ring failure:
(2) Bearing inner ring failure:
(3) Bearing rolling element failure:
(4) Bearing cage failure:
and finally, calculating the error between the peak frequency in the envelope spectrum and the theoretical characteristic frequency to obtain the accuracy of the fault simulation method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310087958.8A CN116011112A (en) | 2023-01-29 | 2023-01-29 | Aeroengine bearing fault data simulation method based on transmission characteristics of casing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310087958.8A CN116011112A (en) | 2023-01-29 | 2023-01-29 | Aeroengine bearing fault data simulation method based on transmission characteristics of casing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116011112A true CN116011112A (en) | 2023-04-25 |
Family
ID=86019338
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310087958.8A Pending CN116011112A (en) | 2023-01-29 | 2023-01-29 | Aeroengine bearing fault data simulation method based on transmission characteristics of casing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116011112A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116380444A (en) * | 2023-06-05 | 2023-07-04 | 滨州鲁德曲轴有限责任公司 | Fault sound data processing system and processing method for crankshaft fault analysis |
-
2023
- 2023-01-29 CN CN202310087958.8A patent/CN116011112A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116380444A (en) * | 2023-06-05 | 2023-07-04 | 滨州鲁德曲轴有限责任公司 | Fault sound data processing system and processing method for crankshaft fault analysis |
| CN116380444B (en) * | 2023-06-05 | 2023-08-25 | 滨州鲁德曲轴有限责任公司 | Fault sound data processing system and processing method for crankshaft fault analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Han et al. | Fault feature extraction of low speed roller bearing based on Teager energy operator and CEEMD | |
| CN106408088B (en) | A kind of rotating machinery method for diagnosing faults based on deep learning theory | |
| Bin et al. | Early fault diagnosis of rotating machinery based on wavelet packets—Empirical mode decomposition feature extraction and neural network | |
| CN105424359B (en) | A kind of gear and bearing mixed fault feature extracting method based on sparse decomposition | |
| CN105806613A (en) | Planetary gear case fault diagnosis method based on order complexity | |
| CN112629840A (en) | Aircraft engine double-rotor-support-casing tester and testing method thereof | |
| CN112461934B (en) | Aero-engine blade crack source positioning method based on acoustic emission | |
| CN108919116B (en) | Ocean current generator unbalanced stator current fault diagnosis method based on MCCKAF-FFT-Softmax | |
| CN116011112A (en) | Aeroengine bearing fault data simulation method based on transmission characteristics of casing | |
| CN103018044A (en) | Bearing combined failure diagnosis method based on improved impact dictionary matching pursuit | |
| CN106055734A (en) | Method for failure diagnosis of wind power gearbox | |
| CN108036940A (en) | A kind of Method for Bearing Fault Diagnosis | |
| CN105352726B (en) | A kind of method for diagnosing faults of gear | |
| CN114813123A (en) | Rolling bearing weak fault diagnosis method based on PSO-VMD-MCKD | |
| Sun et al. | Application of a novel improved adaptive CYCBD method in gearbox compound fault diagnosis | |
| Amin et al. | Wind turbine gearbox fault diagnosis using cyclostationary analysis and interpretable CNN | |
| CN111649855A (en) | Dynamometer signal acquisition method based on compressive sensing | |
| CN106951654A (en) | Complicated coupling structural dynamic characteristics parameter recognition methods based on bounded-but-unknown uncertainty | |
| Liang et al. | Dynamic Simulation of Tooth Root Crack Failure in Planetary Gearbox Based on Torsional Vibration Signal | |
| Hou et al. | Feature Extraction of Weak‐Bearing Faults Based on Laplace Wavelet and Orthogonal Matching Pursuit | |
| Xia et al. | Dynamic modeling of gearbox based on virtual-physical interaction | |
| Liu et al. | Study on Fault Diagnosis Method of Key Components of the Gearbox Under Variable Working Conditions Based on Improved VMD Algorithm | |
| Xiao et al. | Gas turbine blade passing frequency reconstruction and its application for blade fracturing fault diagnosis | |
| Howard | Vibration signal processing using MATLAB | |
| Ye et al. | Analysis of Parallel Misalignment of Gear Coupling in Rotor System Using EEMD-median Filter Method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |