CN-120445660-B - Synchronous vibration acoustic identification method, system, medium and equipment for aero-fan blades

CN120445660BCN 120445660 BCN120445660 BCN 120445660BCN-120445660-B

Abstract

A synchronous vibration acoustic identification method, a system, a medium and equipment for a fan blade of an aero-fanning fan comprise the steps of determining a microphone array layout through parameters of a fan model, synchronously collecting multichannel sound pressure time domain signals of the fan through the microphone array, constructing a time domain signal matrix according to sequence arrangement of the multichannel sound pressure time domain signals, carrying out frequency spectrum analysis on the sound pressure time domain signals by adopting a fast Fourier transform algorithm to obtain a spectrogram of the sound pressure time domain signals of each channel, monitoring whether the amplification of the amplitude of the passing frequency of the blade of the fan exceeds a first preset value or not through the spectrogram, and if broadband noise components with the frequency band distribution larger than a preset range do not exist, indicating that the fan is in a normal working state, reflecting the vibration of the blade through frequency and the sound mode of frequency multiplication of the blade when the fan blade synchronously vibrates, and outputting the vibration pitch diameter and the frequency of a fan blade disc through the sound mode of the blade through the frequency based on the frequency coupling relation.

Inventors

QIAO BAIJIE
WEN BI
ZHAO RUI
Li Zepi
WANG YANAN
LI XINGJIAN
CHEN XUEFENG

Assignees

西安交通大学
太行国家实验室

Dates

Publication Date: 20260512
Application Date: 20250328

Claims (9)

1. A synchronous vibration acoustic identification method for a fan blade of an aero-engine is characterized by comprising the following steps: In a first step (S1), determining a microphone array layout by parameters of a fan model; In a second step (S2), the number of microphones is used Mounting angle of microphone The method comprises the steps of constructing a microphone array, synchronously collecting sound pressure time domain signals of multiple channels of a fan through the microphone array, and constructing a time domain signal matrix for the sound pressure time domain signals of the multiple channels according to sequence arrangement; In the third step (S3), a fast Fourier transform algorithm is adopted to carry out spectrum analysis on the sound pressure time domain signal to obtain a spectrogram of the sound pressure time domain signal of each channel, whether broadband noise components with the amplitude of the blade passing frequency of the fan exceeding a first preset value and the distribution of the occurrence frequency bands being larger than a preset range are generated or not is monitored through the spectrogram, if no broadband noise components which indicate that the fan is in a normal working state are generated, if the amplitude of the blade passing frequency of the fan exceeds the first preset value and the distribution of the occurrence frequency bands being larger than the preset range are generated, the fourth step (S4) is carried out; In the fourth step (S4), an acoustic mode spectrogram of the blade passing frequency is obtained through a single-frequency acoustic mode decomposition method, whether the fan blade synchronously vibrates is judged by monitoring whether the blade synchronous vibration characteristic of the fan appears in the spectrogram, and when the fan blade synchronously vibrates, the change of the amplitude of the conversion static interference mode of the fan blade appears in the acoustic mode spectrogram of the blade passing frequency exceeds a second preset value; In the fifth step (S5), a spectrum chart of a sound pressure time domain signal is obtained through continuous wideband acoustic mode decomposition, the abscissa of the spectrum chart represents a frequency analysis range, the ordinate represents the acoustic mode monitoring range, the depth of color represents the amplitude, the deeper and larger the color, by monitoring whether the synchronous vibration characteristics of the fan blade occur in the spectrum chart, when the synchronous vibration occurs in the fan blade, the broadband noise mode amplitude in the spectrum chart is amplified by more than a third preset value, and the change of the amplitude of the rotating and static interference mode at the passing frequency of the blade exceeds the second preset value and propagates to the peripheral frequency, and is combined with the passing frequency sound source of the blade to form a cross bright band phenomenon; In the sixth step (S6), when the fan blade synchronously vibrates, the blade passing frequency and the acoustic mode of the frequency multiplication reflect the vibration of the blade, and at this time, the vibration frequency of the blade disc is coupled with the blade passing frequency, and the vibration pitch diameter and the frequency of the fan blade disc are output through the acoustic mode of the blade passing frequency based on the frequency coupling relation.
2. The synchronous vibroacoustic identification method of a fan blade according to claim 1, wherein the second step (S2) comprises the steps of: S201, measuring sound pressure signals of the fan by using the annular sound array, wherein the measured sound pressure time domain signals are Wherein the length of the time domain signal sequence measured by the single microphone is Subscript of Microphones corresponding to the installation angle positions respectively; S202, constructing a time domain signal matrix according to sound pressure time domain signals measured by microphones at different installation angle positions Wherein the elements are Representation of Measured by microphones corresponding to the angular positions of the mounting The signal is transmitted to the host computer via the communication network, When a uniform acoustic array layout is used, a time domain signal matrix Is of the size of When the non-uniform acoustic array layout scheme with few measuring points is adopted, the time domain signal matrix Is of the size of 。
3. The method for synchronous vibroacoustic identification of a fan blade according to claim 2, wherein the third step (S3) comprises, S301, time domain signal matrix Each column of (2) is fourier transformed to obtain a frequency domain matrix Wherein the elements are Representation of The signals measured by the microphones corresponding to the mounting angle positions The amplitude at the frequency is such that, Length of (2) According to the nyquist sampling law, , In a uniform acoustic array layout, frequency domain matrices Is of the size of Frequency domain matrix in non-uniform acoustic array layout Is of the size of , S302, according to the frequency domain matrix Drawing frequency spectrograms of signals measured by sensors at different mounting angles, and passing frequencies of blades of fans The amplification of the amplitude exceeds a first predetermined value and a broadband noise component having a frequency band distribution greater than a predetermined range occurs.
4. A synchronous vibroacoustic identification method for a fan blade according to claim 3, wherein the fourth step (S4) comprises, S401, subscript when uniform acoustic array layout is adopted Microphone corresponding to installation angle position at preset frequency Frequency domain signal at Seen as a linear superposition of different circumferential acoustic modes, i.e. Constructing a transformation matrix The form is as follows: , when the non-uniform acoustic array layout with few measuring points is adopted, an observation matrix is constructed according to the randomly selected sensor installation angles Observation matrix The size is as follows , , S402, when the uniform acoustic array layout is adopted, the frequency domain matrix is formed Performing space Fourier transform to obtain wave number domain matrix , Wherein Is a frequency domain matrix Is used to determine the transposed matrix of (a), Wherein Representing a transformation matrix Is a pseudo-inverse of (a), When the non-uniform acoustic array layout with few measuring points is adopted, the compressed sensing model is as follows Sparse dictionary Is formed by an orthogonal fourier transform basis, Sparse reconstruction of wavenumber domain matrix based on compressed sensing model for sensing matrix ; S403, according to the obtained wave number domain matrix When the fan blade synchronously vibrates, the excitation frequency of the sound source is coupled with the vibration frequency of the blade, and the change of the amplitude of the rotating and static interference mode of the sound source appears in the sound mode spectrogram of the passing frequency of the blade to exceed a second preset value.
5. The method for synchronous vibroacoustic identification of a fan blade according to claim 4, wherein the fifth step (S5) comprises, S501, the wave number domain matrix Expanding to the full frequency domain and wavenumber domain matrix The size is expanded into Wherein the elements are Expressed in frequency Location No The amplitude of the acoustic mode of the order, , S502, according to the wave number domain matrix Drawing a spectrogram, observing that the broadband noise modal amplitude in the spectrogram is amplified by more than a third preset value, and the change of the rotating and static interference modal amplitude at the passing frequency of the blade is more than the second preset value and propagates to the peripheral frequency, and combining the blade and the passing frequency sound source into a cross bright band phenomenon to define as the synchronous vibration voiceprint identification characteristic of the fan blade, thereby judging whether the synchronous vibration of the fan blade occurs.
6. The method of claim 1, wherein in the sixth step (S6), the vibration pitch diameter and frequency of the fan blade disk are outputted when the fan blade vibrates synchronously, S601, when the fan blades synchronously vibrate, the vibration frequency of the blade disc is coupled with the blade passing frequency, the acoustic mode order is equal to the pitch diameter number of the blade, and the acoustic source excitation frequency is the same With the vibration frequency of the blade Is not limited to the above-mentioned ones, In which, in the process, Indicating the rotation frequency of the rotor, S602, calculating the vibration frequency and the vibration pitch diameter of the blade according to the excitation frequency and the modal order of the sound source in the spectrogram, and outputting the calculated vibration frequency and vibration pitch diameter.
7. A fan blade synchronous vibration identification system implementing the method of any of claims 1-6, comprising: the sound field measuring module comprises a sound array measuring submodule and a data acquisition submodule and is used for measuring sound field information transmitted to a sound array installation position in a pipeline when the fan works; The frequency spectrum analysis module is used for transforming the time domain sound field signal at the position of the sound array into a frequency domain and detecting whether abnormal change of the passing frequency of the blade and abnormal frequency components exist or not; the sound mode decomposition module is used for carrying out single sound mode decomposition and continuous broadband sound mode decomposition, transforming sound field information from a frequency domain to a wave number domain, and detecting whether synchronous vibration characteristics of the fan blades appear in a sound mode spectrogram and a spectrum chart; and the synchronous vibration calculation module is used for judging whether the fan blades synchronously vibrate or not and outputting the vibration pitch diameter and frequency of the fan blade disc.
8. A computer storage medium comprising computer instructions which, when run on a computer, cause the computer to perform the method of any of claims 1-6.
9. An electronic device, the electronic device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, The processor, when executing the program, implements the method of any one of claims 1-6.

Description

Synchronous vibration acoustic identification method, system, medium and equipment for aero-fan blades Technical Field The invention relates to the technical field of fan noise testing, in particular to a synchronous vibration acoustic identification method, a synchronous vibration acoustic identification system, a synchronous vibration acoustic identification medium and synchronous vibration acoustic identification equipment for a blade of an aero-engine fan. Background Fans are typically multi-physical field highly coupled systems, with internal blades subjected to a combination of structural elastic forces, rotational inertial forces, and unsteady flow excitations. When the fan works under abnormal conditions, the key thin-wall structure inside the fan induces abnormal vibration under the interaction of high centrifugal load, vibration alternating load and pneumatic excitation load. On one hand, certain unsteady flow excitation can lead the vibration amplitude of the blade to increase at an exponential speed, so that the blade is forced to generate high-cycle fatigue failure, form cracks and even cause fatigue fracture of the blade, on the other hand, abnormal vibration of the fan can change an internal flow field structure to threaten the pneumatic stability of a compression system, on the other hand, abnormal vibration of a key thin-wall structure forms a new pneumatic noise source, and under the specific low Mach number condition, sound waves can act on the structure to influence the vibration state. At present, a strain gauge is usually installed on a rotating blade to monitor the vibration state of the fan blade, so that the cost is high, and an invasive measurement mode can generate interference on a fan flow field. The acoustic sensor has the characteristics of high sensitivity, short transmission path and the like, is used for non-invasive measurement, and does not interfere with a fan flow field. The key information such as the vibration frequency and pitch diameter of the blade can be fully reflected by carrying out modal decomposition on the acoustic array signals, the diagnosis precision of a typical abnormal state is improved by the fan blade vibration monitoring method driven by acoustic characteristics, and a new thought is provided for monitoring the fan state of the aeroengine. The large-amplitude synchronous vibration of the fan blades not only can increase the noise of the fan, but also can cause serious fatigue failure of the blades, even cause damage and fracture of the blades, and seriously threaten the safe and stable operation of the high-performance aeroengine fan. The accurate identification and analysis of their fault signature is therefore critical for the maintenance and repair of aeroengines. Conventional fault detection methods mainly rely on visual inspection, nondestructive detection technology, vibration analysis and other means, however, these methods often have limitations in facing synchronous vibration faults of fan blades. The above information disclosed in the background section is only for enhancement of understanding of the background of the invention and therefore may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Disclosure of Invention The invention provides a sound signal measuring method, a sound signal measuring system, a sound signal measuring medium and sound signal processing method, wherein the sound signal measuring medium is used for realizing the conversion of a sound field signal time domain, a sound frequency domain and a wave number domain, so that a more effective sound pattern diagram of a fan blade synchronous vibration detection index is obtained, and the fault characteristics of the sound signal in the fan blade synchronous vibration are identified. The synchronous vibration acoustic identification method for the aero-fan blade comprises the following steps: in the first step, determining a microphone array layout through parameters of a fan model; In the second step, the number of microphones is used Mounting angle of microphoneThe method comprises the steps of constructing a microphone array, synchronously collecting sound pressure time domain signals of multiple channels of a fan through the microphone array, and constructing a time domain signal matrix for the sound pressure time domain signals of the multiple channels according to sequence arrangement; in the third step, a fast Fourier transform algorithm is adopted to carry out spectrum analysis on the sound pressure time domain signal, a spectrogram of the sound pressure time domain signal of each channel is obtained, whether the amplification of the blade passing frequency amplitude of the fan exceeds a first preset value or not is monitored through the spectrogram, and broadband noise components with the occurrence frequency band distribution larger than a preset range are generated, if no broadband noise component