Vibration prediction method for accelerated rotating aero-engine blades under aerodynamic and thermal shock

By combining a feedforward neural network and a harmonic aerodynamic load model, a dynamic motion differential equation was established, which solved the problem of vibration prediction of accelerated rotating aero-engine blades under unsteady aerodynamic loads and thermal shocks, and achieved accurate blade vibration prediction and resonance type identification.

CN123116505APending Publication Date: 2026-11-17NORTHEASTERN UNIV CHINA +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611593035.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-15
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies struggle to accurately predict the vibration response of accelerating rotating aero-engine blades under unsteady aerodynamic loads and thermal shocks. In particular, under the combined effects of rotational-static interference, functionally graded materials, and thermal shock, existing models cannot accurately describe the time-varying characteristics of unsteady aerodynamic loads and the time-varying structural properties caused by thermal shocks, resulting in insufficient vibration prediction accuracy.

Method used

By combining a feedforward neural network with a harmonic aerodynamic load model, a dynamic motion differential equation is established. Through a three-dimensional transient computational fluid dynamics model and a neural network model, the unsteady aerodynamic load and thermal shock effects on the blade during accelerated rotation are described. The blade kinetic energy and potential energy models are constructed, the dynamic motion differential equation is derived, and the resonance peaks and frequency characteristics are identified.

Benefits of technology

It enables accurate vibration prediction of accelerated rotating aero-engine blades under unsteady aerodynamic loads and thermal shocks, improves the reconstruction accuracy of unsteady aerodynamic loads with rotational-static interference, and can accurately predict blade dynamic response and resonance type, thus improving prediction accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123116505A_ABST
    Figure CN123116505A_ABST
Patent Text Reader

Abstract

This invention provides a method for predicting the vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, relating to the field of mechanical engineering technology. It employs a combination of a feedforward neural network and a harmonic aerodynamic load model to obtain the unsteady aerodynamic loads throughout the blade's entire motion time domain, establishes the blade's transient temperature field, and constructs the blade's dynamic motion differential equations. This enables the prediction of the blade's vibration response under the combined effects of unsteady aerodynamic loads caused by rotational-static interference, accelerated rotation, thermal shock, and functionally graded material properties. This invention can achieve vibration response prediction and resonance determination of blades under complex operating conditions.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of mechanical engineering technology, and in particular to a method for predicting the vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock. Background Technology

[0002] Rotating blades are one of the key components of aero-engines, simultaneously subjected to centrifugal force, aerodynamic loads, and thermal loads during service. Among these, rotation-stationary interference is one of the main sources of unsteady aerodynamic loads on blades, significantly affecting their vibration response. Under unsteady conditions such as aero-engine startup and acceleration, blade speed, temperature field, and aerodynamic loads all change rapidly over time, significantly altering the blade's dynamic response characteristics. Therefore, researching vibration prediction of accelerating rotating aero-engine blades under unsteady aerodynamic loads and thermal shocks has significant engineering implications.

[0003] Rotor-to-stationary interference is a common unsteady flow phenomenon in the rotor-to-stationary blade arrays of aero-engines. The relative motion between the rotor-to-stationary blade arrays creates rotor-to-stationary interference, causing periodic pressure fluctuations on the rotor blade surface, which in turn generates unsteady aerodynamic loads with the blade passage frequency and its harmonics as the main frequency components. During accelerated rotation, the blade passage frequency is not constant but changes continuously with the rotational speed. If a load model with a fixed frequency, fixed amplitude, or fixed phase at a constant rotational speed is still used, it is difficult to accurately describe the time-varying characteristics of the aerodynamic loads during actual acceleration.

[0004] Meanwhile, with the continuous increase in thrust-to-weight ratio and turbine inlet temperature of aero-engines, the service environment of blades is becoming increasingly harsh, and functionally graded material (FJT) blades are gradually attracting attention. FJTs are typically composed of a continuous transition between metallic and ceramic phases according to a certain pattern, and their material parameters vary with spatial location and ambient temperature. During engine start-up and acceleration, blades often experience severe thermal shock, and the temperature field evolves rapidly over time. Material parameters and thermal expansion deformation change with temperature, thus affecting the blade's natural frequency and forced vibration response.

[0005] Existing aerodynamic load models caused by static-rotation interference are mostly established for constant speed conditions, and usually represent aerodynamic loads as harmonic functions with constant amplitudes, which makes it difficult to reflect the time-varying evolution of frequency, amplitude, and phase during accelerated rotation. Some studies on accelerated conditions only use linear amplitude correction or constant amplitude approximation, which cannot accurately describe the nonlinear changes of unsteady aerodynamic loads during acceleration, resulting in insufficient vibration prediction accuracy. Existing blade dynamics models are mostly focused on room temperature or steady-state temperature fields, and lack unified consideration of transient temperature changes caused by thermal shock, temperature-dependent characteristics of functionally graded materials, and centrifugal stiffening and rotational softening effects during accelerated rotation. Existing methods do not adequately consider the influence of time-varying unsteady aerodynamic loads and time-varying structural characteristics caused by thermal shock on blade dynamic response during accelerated rotation, making it difficult to accurately predict the occurrence time and peak amplitude of bending resonance response. Summary of the Invention

[0006] In view of at least one of the above technical problems, this application provides a method for predicting the vibration of an aero-engine blade under aerodynamic and thermal shock. The method uses a combination of a feedforward neural network and a harmonic aerodynamic load model to obtain the unsteady aerodynamic load in the entire motion time domain of the blade and establishes a more complete dynamic motion differential equation. The aim is to achieve the prediction of blade vibration response that simultaneously considers the unsteady aerodynamic load caused by rotation-to-stationary interference, the accelerated rotation process, thermal shock and functionally graded material properties.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0008] A method for predicting the vibration of an accelerated rotating aero-engine blade under aerodynamic and thermal shock includes the following steps:

[0009] A three-dimensional transient computational fluid dynamics model was established, and a harmonic aerodynamic load model under constant rotational speed was also established.

[0010] Introducing rotor blade rotation angle Establish an expression for aerodynamic loads under accelerated rotation conditions; calculate the r-th harmonic of the instantaneous blade passing frequency and its rate of change based on the expression for aerodynamic loads under accelerated rotation conditions.

[0011] A neural network aerodynamic load model under accelerated rotation conditions is established, and the neural network aerodynamic load model under accelerated rotation conditions is combined with the harmonic aerodynamic load model under constant speed to obtain the unsteady aerodynamic load model of the blade in the entire motion time domain.

[0012] Construct the blade kinetic energy, the total blade potential energy, and the work done by the unsteady aerodynamic loads throughout the entire blade motion time domain;

[0013] Based on the blade kinetic energy, the total blade potential energy, and the work done by the unsteady aerodynamic loads in the entire blade motion time domain, the dynamic differential equation of the blade under unsteady aerodynamic loads and thermal shock is obtained.

[0014] Based on the differential equations of dynamic motion of the blades under unsteady aerodynamic loads and thermal shocks, the radial displacement and bending displacement of the rotor blades throughout the analysis time are obtained; resonance peaks are identified from the time-domain response of the bending displacement, thereby determining the excitation frequency order, natural frequency order, and resonance type of the rotor blades.

[0015] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the three-dimensional transient computational fluid dynamics model includes the stator blade flow channel, the rotor blade flow channel, and the rotor-stator interface.

[0016] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the establishment of a three-dimensional transient computational fluid dynamics model and a harmonic aerodynamic load model under constant rotational speed include:

[0017] Mesh refinement is performed on the surface of the stator and rotor blades and their adjacent areas. The rotational motion of the rotor relative to the stator is simulated by the sliding mesh method. The rotor blade speed is set by a user-defined function to establish a three-dimensional transient computational fluid dynamics model.

[0018] A harmonic aerodynamic load model under constant rotational speed is established as follows:

[0019] ;

[0020] in, The aerodynamic load acting on the surface of the rotor blades; This is a constant term for the area-weighted average pressure. Let be the amplitude of the r-th harmonic component; The phase of the r-th harmonic component; This indicates the number of stator blades located upstream of the rotor blades. The value represents the rotor blade speed, in rad / s; r is the harmonic order; and t is time.

[0021] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the establishment of the aerodynamic load expression under accelerated rotation conditions includes:

[0022] ;

[0023] in, .

[0024] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the calculation of the r-th harmonic of the instantaneous blade passage frequency and its rate of change based on the aerodynamic load expression under accelerated rotation conditions includes:

[0025] In the expression for aerodynamic loads under accelerated rotation conditions, the instantaneous phase of the r-th harmonic component ,as follows:

[0026] ;

[0027] Based on the relationship between instantaneous frequency and instantaneous phase, the r-th harmonic of the instantaneous blade passage frequency... ,as follows:

[0028] ;

[0029] When r=1, for the accelerated rotation process with constant angular acceleration β, we have Derivation of the instantaneous blade passing frequency when r=1 and its rate of change The expressions are as follows:

[0030] ;

[0031] ;

[0032] Calculate the r-th harmonic of the instantaneous blade passing frequency. and its rate of change They are respectively:

[0033] ;

[0034] .

[0035] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the establishment of a neural network aerodynamic load model under accelerated rotation conditions includes:

[0036] Training samples are constructed using aerodynamic loads at various time points obtained from a three-dimensional transient computational fluid dynamics model, and the input feature vectors at the corresponding time points are used as input. Using the aerodynamic load at the corresponding moment as input and the aerodynamic load at the corresponding moment as output, a feedforward neural network is trained to establish a neural network aerodynamic load model under accelerated rotation conditions. ,in, This represents a nonlinear mapping relationship constructed using a feedforward neural network, with the input feature vector... Defined as:

[0037] ;

[0038] The superscript T indicates vector transpose.

[0039] Furthermore, according to the aforementioned method for predicting the vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the construction of blade kinetic energy and total blade potential energy includes:

[0040] Let the position vector of any point P on the blade in the global coordinate system be denoted as... Thus constructing the blade kinetic energy ,as follows:

[0041] ;

[0042] In the formula, For the volume of the blade, For mass density, Let be the velocity vector of any point on the blade in the global coordinate system;

[0043] Considering tensile, bending, shear deformation and thermal expansion effects, the blade strain energy Represented as:

[0044] ;

[0045] in, Here, b is the shear correction factor, h is the blade width, and L is the blade length. and These are normal stress and shear stress, respectively. For positive strain, For shear strain, The effective linear thermal expansion coefficient for functionally graded materials; For the temperature rise, , This represents the transient temperature field of the rotor blades. This refers to the initial temperature of the blade. y and z are the coordinates of the blade's length direction, thickness direction, and width direction, respectively;

[0046] Potential energy generated by centrifugal stiffening Represented as:

[0047] ;

[0048] in, This represents the centrifugal force exerted on the cross-section at position x of the blade. This represents the lateral displacement of the blade. for First-order partial derivative with respect to coordinate x;

[0049] Depend on and Constructing the total potential energy of the blade ,Right now .

[0050] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the work done by the unsteady aerodynamic loads throughout the entire motion time domain of the blade is as follows:

[0051]

[0052] In the formula, The work done by the unsteady aerodynamic loads over the entire time domain of the blade's motion.

[0053] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the differential equations of motion of the blade under unsteady aerodynamic loads and thermal shock are obtained based on the blade's kinetic energy, total potential energy, and the work done by the unsteady aerodynamic loads throughout the blade's entire motion time domain. These equations include:

[0054] The differential equations of motion for the rotor blades are derived from Hamilton's principle as follows:

[0055] ;

[0056] In the formula, The variational function is represented by t1 and t2, which are the start and end times of the analysis time interval, respectively. The blade kinetic energy, total blade potential energy, and the work done by the unsteady aerodynamic loads throughout the blade's entire motion time domain are substituted into the equation. Considering structural proportional damping, the differential equations of motion for the accelerated rotating aero-engine blades under unsteady aerodynamic loads and thermal shocks can be obtained:

[0057] ;

[0058] In the formula, , and These are the blade's generalized coordinate vector, generalized velocity vector, and generalized acceleration vector, respectively. , , , , , and These represent the blade mass matrix, gyroscope matrix, damping matrix, structural stiffness matrix, centrifugal stress stiffening matrix, rotational softening matrix, and stiffness matrix caused by angular acceleration, respectively. , and These are the generalized force vectors caused by rotational effects, thermal expansion effects, and aerodynamic loads, respectively.

[0059] Furthermore, according to the aforementioned method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, the step of identifying resonance peaks from the time-domain response of bending displacement to determine the excitation frequency order, natural frequency order, and resonance type of the rotor blade includes:

[0060] Identify resonance peaks and determine their occurrence times from the time-domain response of bending displacement;

[0061] The response frequency at the resonance peak is extracted from the three-dimensional spectrum of the time-domain response to bending displacement.

[0062] The response frequency at the resonance peak moment is matched with the instantaneous blade passage frequency and its harmonics, as well as the time-varying natural frequencies of each order, to determine the excitation frequency order, natural frequency order, and resonance type of the rotor blade.

[0063] The beneficial effects of adopting the above technical solution are as follows:

[0064] (1) The present invention provides a method for predicting the vibration of aero-engine blades under aerodynamic and thermal shock. By correcting the expression of the rth harmonic of the instantaneous blade passing frequency and its rate of change through the rotor blade rotation angle, it can accurately describe the time-varying law of aerodynamic load caused by rotor-stationary interference during accelerated rotation.

[0065] (2) The present invention uses a feedforward neural network to model the unsteady aerodynamic load caused by the rotation-to-station interference under accelerated rotation conditions. The nonlinear mapping capability of the feedforward neural network is used to describe the nonlinear change of the aerodynamic load over time during the acceleration phase. After the rotor blade reaches a constant speed, the aerodynamic load is extended using a harmonic aerodynamic load model. The two models are combined to obtain the unsteady aerodynamic load model in the entire motion time domain of the blade, thereby improving the reconstruction accuracy of the unsteady aerodynamic load caused by the rotation-to-station interference.

[0066] (3) The present invention establishes a blade dynamic motion differential equation that simultaneously considers thermal shock, temperature correlation of material parameters and functional gradient distribution characteristics, as well as rotational dynamic effects of centrifugal stiffening and rotational softening, which can more realistically predict the dynamic response of rotor blades.

[0067] (4) The present invention identifies the time of resonance peak occurrence by bending displacement time domain response and obtains the response frequency of resonance peak time based on the three-dimensional spectrum of bending displacement time domain response; further, the response frequency of resonance peak time is matched with the instantaneous blade passing frequency and its harmonics, and the time-varying natural frequencies of each order, thereby determining the excitation frequency order, natural frequency order and resonance type of rotor blade. Attached Figure Description

[0068] Figure 1 This is a flowchart of a method for predicting vibration of an accelerated rotating aero-engine blade under aerodynamic and thermal shock, provided in an embodiment of the present invention.

[0069] Figure 2 This is a schematic diagram of the rotating and stationary blade array and functionally graded blades under load provided in an embodiment of the present invention; Figure 2 (a) is a schematic diagram of a rotating-stationary bladed array with flow disturbances caused by the interference between the rotating and stationary phases; Figure 2 (b) is a model of an accelerated rotating functional gradient blade under unsteady aerodynamic loads and thermal shock;

[0070] Figure 3 This is a schematic diagram of a computational fluid dynamics model for obtaining unsteady aerodynamic loads caused by the interference between static and dynamic states, provided in an embodiment of the present invention. Figure 3 (a) is the full-channel geometric model; Figure 3 (b) is a single-channel periodicity calculation model; Figure 3 (c) shows the mesh distribution of the mid-section along the blade length in a single-channel computational domain;

[0071] Figure 4 This is a time-domain and time-frequency characteristic analysis diagram of aerodynamic loads under accelerated rotation conditions provided in an embodiment of the present invention; Figure 4 (a) is the time-domain curve of aerodynamic load; Figure 4 (b) is a time-frequency distribution diagram of aerodynamic loads; Figure 4 (c) shows the instantaneous dominant frequency ridge and its linear fitting results;

[0072] Figure 5 The fitting results of different aerodynamic load modeling schemes during the accelerated rotation stage provided in the embodiments of the present invention are shown in the figure. Figure 5 (a) is a linear amplitude correction scheme; Figure 5 (b) is a linear amplitude-phase correction scheme; Figure 5 (c) is the modeling scheme for the feedforward neural network;

[0073] Figure 6 The following are diagrams showing the aerodynamic load reconstruction results under different operating conditions within an 8-second analysis time, as provided in the embodiments of the present invention. Figure 6 (a) The corresponding working condition is β=200rad / s², η=30°; Figure 6 (b) For the operating condition β=200rad / s², η=15°; Figure 6 (c) Corresponding to the working condition of β=100rad / s², η=30°;

[0074] Figure 7 A comparison diagram of blade tip vibration response (M=+∞) provided for verifying the correctness of the model in an embodiment of the present invention; Figure 7 (a) Comparison of the radial displacement of the blade tip predicted by the theoretical model and the finite element model; Figure 7(b) is a comparison diagram of the blade tip bending displacement predicted by the theoretical model and the finite element model; Figure 7 (c) is a waterfall plot of the bending displacement spectrum during the acceleration process;

[0075] Figure 8 This is a time-domain diagram of the blade tip radial displacement provided in an embodiment of the present invention; Figure 8 (a) is the time-domain diagram of the tip radial displacement when the angular acceleration β = 0 rad / s²; Figure 8 (b) is the time-domain plot of the tip radial displacement when the angular acceleration β = 100 rad / s²; Figure 8 (c) is the time-domain plot of the tip radial displacement when the angular acceleration β = 200 rad / s²;

[0076] Figure 9 This is a time-domain diagram of blade tip bending displacement provided in an embodiment of the present invention; Figure 9 (a) is the time-domain plot of the tip bending displacement when the angular acceleration β = 0 rad / s²; Figure 9 (b) is the time-domain plot of the tip bending displacement when the angular acceleration β = 100 rad / s²; Figure 9 (c) is the time-domain plot of the tip bending displacement when the angular acceleration β = 200 rad / s²;

[0077] Figure 10 A waterfall plot of bending displacement spectrum under different angular acceleration conditions provided in an embodiment of the present invention; Figure 10 (a) is a waterfall plot of the bending displacement spectrum when the angular acceleration β = 0 rad / s²; Figure 10 (b) is a waterfall plot of the bending displacement spectrum when the angular acceleration β = 100 rad / s²; Figure 10 (c) is a waterfall plot of the bending displacement spectrum when the angular acceleration β = 200 rad / s²;

[0078] Figure 11 A graph showing the relationship between the first three time-varying natural frequencies and the instantaneous blade passing frequency and its harmonics, provided for an embodiment of the present invention. Figure 11 (a) is a graph showing the relationship when angular acceleration β = 0 rad / s²; Figure 11 (b) is a graph showing the relationship when the angular acceleration β = 100 rad / s²; Figure 11 (c) is a graph showing the relationship when the angular acceleration β = 200 rad / s². Detailed Implementation

[0079] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0080] This embodiment provides a method for predicting the vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock, such as... Figure 1As shown, it is preferably applied to the vibration response prediction and resonance risk assessment of rotor blades in the rotor stator blade array of aero engines or gas turbines.

[0081] like Figure 2 As shown, the relative motion between the rotor and stationary blade rows generates a rotor-stationary interference effect in the flow field, subjecting the rotor blades to unsteady aerodynamic loads that vary with time. Simultaneously, the rotor blades are subjected to thermal shock during operation. Figure 2 (a) is a schematic diagram of a rotating-stationary bladed array with flow disturbances caused by the interference between the rotating and stationary phases. Figure 2 (b) is a model of an accelerated rotating aero-engine blade under unsteady aerodynamic loads and thermal shock. Since all rotor blades have the same structural form, this embodiment selects one rotor blade as a representative blade, based on... A dynamic model of an accelerated rotating aero-engine blade is established using beam theory. The blade length, width, thickness, and installation angle are denoted as L, b, h, and η, respectively, and the rotor radius is denoted as... The blade material consists of a ceramic phase (zirconia) and a metallic phase (titanium alloy), with a continuous gradient change along the thickness direction.

[0082] Unless otherwise stated, the geometry, materials and operating parameters of the rotor blades used in this embodiment are shown in Table 1.

[0083]

[0084] In Table 1, L represents the blade length; b represents the blade width; and h represents the blade thickness. Indicates the radius of the roulette wheel; M represents the power law exponent; Indicates the relative heat transfer coefficient; T0 represents the initial blade temperature; T1 represents the blade ambient temperature; Ω max β represents the maximum rotational speed of the blade; β represents the angular acceleration of the rotor blade rotation; η represents the blade installation angle; This represents the shear correction factor.

[0085] Step 1: Establish a three-dimensional transient computational fluid dynamics model to obtain the unsteady aerodynamic loads on the rotor blades;

[0086] This embodiment first establishes a system for rotating and stationary bladed arrays, as follows: Figure 3 The three-dimensional transient computational fluid dynamics model shown is illustrated. Figure 3 (a) is the full-channel geometric model. Figure 3 (b) is a single-channel period calculation model. Figure 3(c) shows the mesh distribution of the cross-section along the blade length in the single-channel computational domain. This three-dimensional transient computational fluid dynamics model includes the stator blade flow channel, the rotor blade flow channel, and the rotor-stator interface. To reduce computational cost, this embodiment uses a single-channel periodic computational model based on circumferential periodicity, and refines the mesh on the stator and rotor blade surfaces and their adjacent regions. The rotational motion of the rotor relative to the stator is simulated using a sliding mesh method, and the rotor blade speed is set using a user-defined function.

[0087] The area-weighted pressure on the rotor blade surface is extracted to obtain the unsteady aerodynamic load on the rotor blade, denoted as . Specifically, it can be expressed as:

[0088] ;

[0089] in, This represents the total surface area of ​​the rotor blades. P represents the total number of surface units on the rotor blades. i (t) represents the time t. aerodynamic pressure on the surface unit of each rotor blade For the first The area of ​​each rotor blade surface element; this step is used to obtain the true time-domain characteristics of the unsteady aerodynamic load caused by rotor-stationary interference.

[0090] Step 2: Perform spectral analysis on the unsteady aerodynamic loads of the rotor blades to establish a harmonic aerodynamic load model under constant speed; introduce the rotor blade rotation angle. Establish an expression for aerodynamic loads under accelerated rotation conditions; calculate the r-th harmonic of the instantaneous blade passing frequency and its rate of change based on the expression for aerodynamic loads under accelerated rotation conditions.

[0091] Spectral analysis was performed on the unsteady aerodynamic loads on the rotor blades under constant speed conditions. The results show that the main frequency components of the unsteady aerodynamic loads are the blade passage frequency and its harmonics. Therefore, a harmonic aerodynamic load model (hereinafter referred to as the harmonic model) can be used to describe the aerodynamic loads under constant speed, and the amplitude and phase of each harmonic can be identified through an optimization algorithm. Furthermore, the unsteady aerodynamic loads caused by rotor-stationary interference exhibit significant periodicity; therefore, the equivalent aerodynamic force... It can be approximated as the superposition of multiple harmonic components, as follows:

[0092] ;

[0093] in, For constant terms, and These represent the amplitude and phase of the r-th harmonic component, respectively. This indicates the number of stator blades located upstream of the rotor blades; in this embodiment, it is taken as 36. This represents the rotor blade rotational speed, expressed in rad / s. The r-th harmonic of the frequency through which the blade passes is given by the equation... The value is given, and the unit is Hz.

[0094] In addition, equivalent aerodynamic forces It can be obtained by multiplying the area-weighted average pressure of the rotor blade surface by the blade surface area, that is:

[0095] ;

[0096] Since the surface area of ​​the rotor blades remains constant and They exhibit the same temporal evolution patterns and frequency characteristics. Therefore, this embodiment adopts... Characterizing the time-varying features of aerodynamic loads. Accordingly, Can be written as Using the same harmonic expansion form, a harmonic aerodynamic load model is established under constant rotational speed, as follows:

[0097] ;

[0098] in, This is a constant term for the area-weighted average pressure. Let be the amplitude of the r-th harmonic component.

[0099] For accelerated rotation conditions, directly using the harmonic expression at constant speed will lead to inaccurate phase description. Therefore, in this embodiment, the time term in the harmonic model is corrected to the rotor blade rotation angle. The expression for aerodynamic loads under accelerated rotation conditions is:

[0100] ;

[0101] In the formula, The aerodynamic load acting on the rotor blade surface is t, which is time, and r is the harmonic order. In this embodiment, the first three main harmonic components are taken. and Let be the amplitude and phase of the r-th harmonic component, respectively. The number of stator blades, The rotor blade rotation angle;

[0102] In the expression for aerodynamic loads under accelerated rotation conditions, the instantaneous phase of the r-th harmonic component for:

[0103] ;

[0104] Based on the relationship between instantaneous frequency and instantaneous phase, the r-th harmonic of the instantaneous blade passage frequency... for:

[0105] ;

[0106] When r=1, corresponding to the passing frequency of the blade, for the accelerated rotation process with constant angular acceleration β, we have: Therefore, when r=1, the instantaneous blade passing frequency and its rate of change The expressions are as follows:

[0107] ;

[0108] ;

[0109] Correspondingly, the r-th harmonic of the instantaneous blade passage frequency and its rate of change They are respectively:

[0110] ;

[0111] ;

[0112] Figure 4 Taking the working condition of angular acceleration β = 200 rad / s² and installation angle η = 30° as representative, the analysis results of aerodynamic load time-domain curves, time-frequency distribution, and instantaneous dominant frequency change rate are presented. Figure 4 (a) is the time-domain curve of aerodynamic load. Figure 4 (b) is the time-frequency distribution diagram of aerodynamic load. Figure 4 (c) shows the instantaneous dominant frequency ridge extracted from the time-frequency distribution and its linear fitting results. The slope of the fitted line is used to characterize the instantaneous dominant frequency change rate. Table 2 further summarizes the comparison results between the theoretical instantaneous blade passage frequency change rate and the change rate extracted by time-frequency analysis under three sets of operating conditions. It can be seen that the instantaneous dominant frequency change rate extracted by time-frequency analysis under the three sets of operating conditions has good consistency with the theoretical instantaneous blade passage frequency change rate, and the relative error is small, indicating that this method can accurately describe the time-varying law of the dominant frequency of the rotation-to-stationary interference during accelerated rotation.

[0113]

[0114] Step 3: Construct the input feature vector ,by As input, a neural network aerodynamic load model under accelerated rotation conditions is established; the neural network aerodynamic load model under accelerated rotation conditions is combined with the harmonic aerodynamic load model under constant speed to obtain the unsteady aerodynamic load model of the blade in the entire motion time domain.

[0115] To accurately describe the nonlinear characteristics of the amplitude, phase, and frequency of aerodynamic loads during the blade's accelerated rotation phase, three schemes—linear amplitude correction, linear amplitude-phase correction, and feedforward neural network—were used to fit the unsteady aerodynamic loads during the acceleration phase. The fitting results are as follows: Figure 5 As shown, where, Figure 5 (a) shows the fitting results of the linear magnitude correction scheme. Figure 5 (b) shows the fitting results of the linear amplitude-phase correction scheme. Figure 5 (c) shows the fitting results of the feedforward neural network scheme. The comparison results show that the feedforward neural network can more accurately reconstruct the unsteady aerodynamic load during the accelerated rotation phase. Therefore, in this embodiment, the feedforward neural network is used to establish the neural network aerodynamic load model under accelerated rotation conditions.

[0116] The input-output relationship of the neural network aerodynamic load model under the accelerated rotation condition can be expressed as:

[0117] ;

[0118] In the formula, This represents the nonlinear mapping relationship established by the feedforward neural network, with the input feature vector. Defined as:

[0119] ;

[0120] Among them, the three sets of sine and cosine terms correspond to the instantaneous blade passage frequency and its second and third harmonics, respectively. Based on the aforementioned instantaneous blade passage frequency, the time-varying phase term corresponding to the r-th harmonic is: (r=1,2,3). The time term t is used to characterize the time-varying characteristics of the acceleration process, and the sine and cosine terms are used to characterize the main periodic characteristics of the aerodynamic loads of the rotation-to-station interference. The superscript T indicates vector transpose.

[0121] Using the aerodynamic loads calculated at each moment by the aforementioned three-dimensional transient computational fluid dynamics model, training samples are constructed, and the input feature vectors at the corresponding moments are... As input, aerodynamic load As output, the feedforward neural network is trained to obtain the input feature vector. With unsteady aerodynamic loads Nonlinear mapping relationship between .

[0122] During the accelerated rotation phase, the feedforward neural network is used to reconstruct the unsteady aerodynamic load, that is, a neural network aerodynamic load model under accelerated rotation conditions is adopted: During the constant-speed rotation phase after the rotor blade reaches the preset speed, a harmonic aerodynamic load model under constant speed is used. The neural network aerodynamic load model under accelerated rotation conditions is combined with the harmonic aerodynamic load model under constant speed to obtain the unsteady aerodynamic load model for the entire blade motion time domain. When the analysis terminates earlier than the moment the rotor blade reaches the preset speed, the neural network aerodynamic load model under accelerated rotation conditions is used for the entire analysis time domain. Figure 6 (a) For the operating condition corresponding to β=200rad / s², η=30°, Figure 6 (b) For the operating condition β=200rad / s², η=15°, Figure 6 (c) Corresponding to the working condition of β=100rad / s², η=30°.

[0123] Step 4: Establish the transient temperature field of the blade, and construct the blade kinetic energy and total blade potential energy based on the transient temperature field;

[0124] Under the assumption that the temperature changes only along the thickness direction and there is no internal heat source, the transient temperature field of the rotor blades The following one-dimensional transient heat conduction equation is satisfied:

[0125] ;

[0126] In the formula, The thermal conductivity of the blade is... The specific heat capacity of the blade, denoted as blade mass density, y as blade thickness direction coordinate, and t as time;

[0127] The initial and boundary conditions can be written as:

[0128] ;

[0129] In the formula, This represents the convective heat transfer coefficient of the outer surface of the blade. Indicates the initial temperature of the blade. Let h represent the ambient temperature and h be the blade thickness. This allows us to obtain the temperature distribution of the blade along the thickness and time directions under thermal shock. Furthermore, a dimensionless relative heat transfer coefficient is defined. for:

[0130] ;

[0131] in, It is the relative heat transfer coefficient; h is the blade thickness; This represents the convective heat transfer coefficient of the outer surface of the blade; The thermal conductivity of the blade;

[0132] The blades are made of zirconium dioxide in the thickness direction. It consists of a continuous gradient of titanium (Ti-6Al-4V) and titanium alloy (Ti-6Al-4V). Its temperature-dependent material properties are based on... The model is represented as:

[0133] ;

[0134] Where P represents temperature-related material parameters, which can represent thermal conductivity, specific heat capacity, linear thermal expansion coefficient, Young's modulus, Poisson's ratio, or mass density, and T is temperature. , , , , For temperature coefficients, see Table 3.

[0135]

[0136] In Table 3, the superscript a applies to 300K~1100K; the superscript b applies to 1100K~1300K.

[0137] Assuming the metal volume fraction of functionally graded material blades It follows a power-law distribution along the thickness direction, expressed as:

[0138] ;

[0139] In the formula, is the volume fraction of the metal, and M is the power-law exponent.

[0140] The sum of the volume fractions of the metal and ceramic satisfies:

[0141] ;

[0142] In the formula, and These are the volume fractions of ceramics and metals, respectively.

[0143] According to the mixing law, the thermal conductivity of the blade mass density Specific heat capacity Young's modulus linear thermal expansion coefficient Compared to Poisson They are represented as follows:

[0144]

[0145] In the formula, the left side of the equal sign , , , , and These are the effective thermal conductivity, effective mass density, effective specific heat capacity, effective Young's modulus, effective linear thermal expansion coefficient, and effective Poisson's ratio of the functionally graded materials, respectively. and These represent the volume fractions of the ceramic phase and the metallic phase, respectively; the subscripts c and m represent the ceramic phase and the metallic phase, respectively.

[0146] We select a rotor blade as a representative blade. Since the blade thickness is much smaller than its width, we ignore the chordal displacement and only consider the radial displacement u and the lateral bending displacement. and cross section rotation angle .according to According to beam theory, the strain-displacement relationship of a blade is:

[0147] ;

[0148] In the formula, For positive strain, For shear strain, The coordinates are along the blade length direction. The coordinates are along the thickness direction of the blade.

[0149] The obtained transient temperature field Substituting the temperature-dependent material parameter model of the functionally graded material, and considering the thermal expansion effect caused by temperature changes, the stress-strain relationship of the blade is expressed as:

[0150] ;

[0151] In the formula, and These are normal stress and shear stress, respectively. For Young's modulus, It is the linear thermal expansion coefficient. Shear modulus , Poisson's ratio; For the temperature rise, , This represents the initial temperature of the blade.

[0152] The position vector of any point P on the blade in the global coordinate system is denoted as: Constructing blade kinetic energy The details are as follows:

[0153] ;

[0154] In the formula, For the volume of the blade, For mass density, Let be the velocity vector of any point on the blade in the global coordinate system.

[0155] Among them, considering tensile, bending, shear deformation and thermal expansion effects, the blade strain energy Represented as:

[0156] ;

[0157] in, Here, b is the shear correction factor, h is the blade width, and L is the blade length. and These are normal stress and shear stress, respectively. For positive strain, For shear strain, The effective linear thermal expansion coefficient for functionally graded materials; For the temperature rise, , This represents the transient temperature field of the rotor blades. The initial temperature of the blade is given; x, y, and z are the coordinates of the blade's length, thickness, and width, respectively.

[0158] Potential energy generated by centrifugal stiffening Represented as:

[0159]

[0160] in, This represents the centrifugal force exerted on the cross-section at position x of the blade. This represents the lateral displacement of the blade. for First-order partial derivative with respect to coordinate x;

[0161]

[0162] In the formula, L is the blade length, b is the blade width, and h is the blade thickness. The radius of the roulette wheel, The instantaneous rotational speed of the blade.

[0163] Depend on and Constructing the total potential energy of the blade ,as follows:

[0164] ;

[0165] Step 5: Construct the work done by the unsteady aerodynamic loads in the entire motion domain of the blade. Based on the blade kinetic energy, the total potential energy of the blade, and the work done by the unsteady aerodynamic loads in the entire motion domain of the blade, obtain the dynamic differential equation of the blade under unsteady aerodynamic loads and thermal shock.

[0166] Work done by unsteady aerodynamic loads throughout the entire motion domain of the blade The details are as follows:

[0167]

[0168] In the formula, The width of the blade. These are the coordinates along the length of the blade.

[0169] The differential equation of motion of the rotor blades can be derived using Hamilton's principle, and its equation is as follows:

[0170] ;

[0171] In the formula, Let represent the variation, and t1 and t2 be the start and end times of the analysis time interval, respectively. Thus, the blade kinetic energy, total blade potential energy, and the work done by the unsteady aerodynamic loads throughout the entire blade motion time domain are calculated. Substitute into the above formula Considering structural proportional damping, the differential equations of motion for the accelerated rotating aero-engine blades under unsteady aerodynamic loads and thermal shocks can be obtained:

[0172] ;

[0173] In the formula, , and These are the blade's generalized coordinate vector, generalized velocity vector, and generalized acceleration vector, respectively. , , , , , and These represent the blade mass matrix, gyroscope matrix, damping matrix, structural stiffness matrix, centrifugal stress stiffening matrix, rotational softening matrix, and stiffness matrix caused by angular acceleration, respectively. , and These are the generalized force vectors caused by rotational effects, thermal expansion effects, and aerodynamic loads, respectively.

[0174] Step 6: Based on the dynamic differential equations of the blade under unsteady aerodynamic loads and thermal shock, obtain the radial displacement and bending displacement of the rotor blade throughout the analysis time; identify resonance peaks from the bending displacement time-domain response to determine the excitation frequency order, natural frequency order, and resonance type of the rotor blade.

[0175] The time-varying natural frequencies of the blades are obtained by solving the dynamic differential equations using the complex modal method. The radial and bending displacements of the rotor blades are obtained by solving the dynamic differential equations using the Newmark-β numerical integration method throughout the analysis time.

[0176] The resonance determination process is as follows: First, the resonance peak is identified from the bending displacement time domain response and its occurrence time is determined; then, the response frequency at the resonance peak moment is extracted through the three-dimensional spectrum of the bending displacement time domain response; finally, the response frequency at the resonance peak moment is matched with the instantaneous blade passage frequency and its harmonics, and the time-varying natural frequencies of each order, so as to determine the excitation frequency order, natural frequency order and resonance type of the rotor blade.

[0177] To verify the correctness of the method of this invention, the dynamic model in this embodiment is first degraded to the baseline operating condition in existing literature. In this degraded operating condition, thermal shock is not considered, and the unsteady aerodynamic load is simplified to a constant amplitude first harmonic model. By comparing with the results of existing literature, the tip displacement response and frequency domain characteristics show good consistency, and the relative error of the rotational speed when resonance occurs is small, indicating that the dynamic differential equation of motion and numerical solution method established in this invention are reliable.

[0178] Furthermore, this embodiment uses the finite element method to verify the theoretical model under complete operating conditions. In this finite element verification condition, the power law exponent M of the theoretical model is taken as +∞, at which point the blade degenerates into a homogeneous metal blade; the finite element model correspondingly uses a homogeneous Ti-6Al-4V metal material. Except for the power law exponent, the theoretical model and the finite element model use the same geometric and operating condition parameters. Table 4 shows the comparison results of the first three time-varying natural frequencies predicted by the theoretical model and the finite element model. It can be seen that within the 1–8s analysis interval, the trend of the first three natural frequencies predicted by the theoretical model is consistent with the trend of the finite element results, and the numerical agreement is good. The maximum relative errors of the first, second, and third natural frequencies are 3.25%, 3.01%, and 2.75%, respectively, and the errors of each order generally decrease and tend to stabilize with the increase of time. The above results show that the theoretical model established in this embodiment can accurately predict the first three time-varying natural frequencies of the blade under complete operating conditions, which can provide a reliable basis for subsequent forced vibration response analysis.

[0179]

[0180] like Figure 7 As shown, Figure 7 (a) is a comparison diagram of the radial displacement of the blade tip predicted by the theoretical model and the finite element model. Figure 7 (b) is a comparison diagram of the tip bending displacement predicted by the theoretical model and the finite element model. Figure 7(c) is a waterfall plot of the bending displacement spectrum during acceleration. The theoretical predictions for the tip radial displacement, bending displacement, and bending displacement spectrum waterfall plot are consistent with the finite element results. This result demonstrates that the method of this invention can accurately predict the vibration response of an accelerated rotating functionally graded blade under unsteady aerodynamic loads and thermal shock.

[0181] Based on the finite element verification above, the power law exponent is restored to M=20 as listed in Table 1, while all other parameters remain unchanged except for angular acceleration. This embodiment further analyzes the influence of angular acceleration on the blade vibration response. Figure 8 This is a time-domain diagram of the blade tip radial displacement provided in an embodiment of the present invention. Figure 9 This is a time-domain diagram of blade tip bending displacement provided in an embodiment of the present invention. Figure 10 This is a waterfall plot of the bending displacement spectrum under different angular acceleration conditions provided in an embodiment of the present invention. Figure 11 The following diagrams, provided for embodiments of the present invention, illustrate the relationship between the first three time-varying natural frequencies and the instantaneous blade passing frequency and its harmonics, showing angular accelerations of 0, 100, and 200°. There are three scenarios. Specifically, when the angular acceleration is zero, no obvious first-order bending resonance occurs under the conditions analyzed in this embodiment; when the angular acceleration is greater than zero, as the rotational speed increases, the instantaneous blade passage frequency gradually increases, and a significant first-order bending resonance is induced when it approaches the first time-varying natural frequency. As the angular acceleration increases, the instantaneous blade passage frequency increases faster, the resonance condition is met earlier, and therefore the bending resonance peak appears earlier. This result shows that the method of the present invention can not only obtain the blade displacement response, but also accurately predict the time of resonance occurrence.

[0182] This embodiment demonstrates that the method of the present invention can achieve full-time-domain aerodynamic load modeling, transient temperature field solution, vibration response prediction, and resonance risk assessment for accelerated rotating functionally graded blades. This method is applicable to the dynamic analysis of rotor blades under unsteady aerodynamic loads and thermal shocks caused by rotation-to-stationary interference, and can provide technical support for aero-engine blade structural design, material gradient design, and operational safety assessment.

[0183] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the present invention.

Claims

1. A method for predicting accelerated rotating aeroengine blade vibrations under pneumatic and thermal shock, characterized in that, Includes the following steps: A three-dimensional transient computational fluid dynamics model was established, and a harmonic aerodynamic load model under constant rotational speed was also established. Introducing rotor blade rotation angle , the expression of aerodynamic load under accelerated rotation condition is established. Calculate the r-th harmonic and its rate of change of the instantaneous blade passing frequency based on the aerodynamic load expression under accelerated rotation conditions; A neural network aerodynamic load model under accelerated rotation conditions is established, and the neural network aerodynamic load model under accelerated rotation conditions is combined with the harmonic aerodynamic load model under constant speed to obtain the unsteady aerodynamic load model of the blade in the entire motion time domain. Construct the blade kinetic energy, the total blade potential energy, and the work done by the unsteady aerodynamic loads throughout the entire blade motion time domain; Based on the blade kinetic energy, the total blade potential energy, and the work done by the unsteady aerodynamic loads in the entire blade motion time domain, the dynamic differential equation of the blade under unsteady aerodynamic loads and thermal shock is obtained. Based on the differential equations of dynamic motion of the blades under unsteady aerodynamic loads and thermal shock, the radial displacement and bending displacement of the rotor blades throughout the entire analysis time are obtained. Resonance peaks are identified from the time-domain response of bending displacement, thereby determining the excitation frequency order, natural frequency order, and resonance type of the rotor blades.

2. The method for accelerated rotating aeroengine blade vibration prediction under pneumatic and thermal shock according to claim 1, characterized in that, The three-dimensional transient computational fluid dynamics model includes the stator blade flow channel, the rotor blade flow channel, and the rotor-stationary interface.

3. The method of accelerated rotating aeroengine blade vibration prediction under pneumatic and thermal shock according to claim 2, characterized in that, The establishment of the three-dimensional transient computational fluid dynamics model and the establishment of the harmonic aerodynamic load model under constant rotational speed include: Mesh refinement is performed on the surface of the stator and rotor blades and their adjacent areas. The rotational motion of the rotor relative to the stator is simulated by the sliding mesh method. The rotor blade speed is set by a user-defined function to establish a three-dimensional transient computational fluid dynamics model. A harmonic aerodynamic load model under constant rotational speed is established as follows: ; wherein, is the aerodynamic load acting on the rotor blade surface; is the constant term of the area-weighted mean pressure; is the amplitude of the rth harmonic component; is the phase of the rth harmonic component; denotes the number of stator blades upstream of the rotor blade, denotes the rotor blade rotational speed in rad / s; r is the harmonic order; t is the time.

4. The method of accelerated rotating aeroengine blade vibration prediction under pneumatic and thermal shock according to claim 3, characterized in that, The establishment of the aerodynamic load expression under accelerated rotation conditions includes: ; wherein .

5. The method of accelerated rotating aeroengine blade vibration prediction under pneumatic and thermal shock according to claim 4, characterized in that, The calculation of the r-th harmonic of the instantaneous blade passage frequency and its rate of change based on the aerodynamic load expression under accelerated rotation conditions includes: In the expression of the aerodynamic load under the accelerated rotation condition, the instantaneous phase of the rth harmonic component As follows: ; According to the relationship between the instantaneous frequency and the instantaneous phase, the instantaneous blade passes through the rth harmonic of the frequency As follows: ; For the constant angular acceleration β, the acceleration rotation process has , the expressions of the instantaneous blade passing frequency and its rate of change when r = 1 are derived as follows: ; ; calculating the rth harmonic of the instantaneous blade passing frequency and its rate of change are respectively: ; 。 6. The method for accelerated rotating aeroengine blade vibration prediction under pneumatic and thermal shock according to claim 5, characterized in that, The establishment of the neural network aerodynamic load model under accelerated rotation conditions includes: Training samples are constructed using aerodynamic loads at various time points obtained from a three-dimensional transient computational fluid dynamics model, and the input feature vectors at the corresponding time points are used as input. Using the aerodynamic load at the corresponding moment as input and the aerodynamic load at the corresponding moment as output, a feedforward neural network is trained to establish a neural network aerodynamic load model under accelerated rotation conditions. ,in, This represents a nonlinear mapping relationship constructed using a feedforward neural network, with the input feature vector... Defined as: ; The superscript T indicates vector transpose.

7. The method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock according to claim 1, characterized in that, The construction of blade kinetic energy and total blade potential energy includes: Let the position vector of any point P on the blade in the global coordinate system be denoted as... Thus constructing the blade kinetic energy ,as follows: ; In the formula, For the volume of the blade, For mass density, Let be the velocity vector of any point on the blade in the global coordinate system; Considering tensile, bending, shear deformation and thermal expansion effects, the blade strain energy Represented as: ; in, Here, b is the shear correction factor, h is the blade width, and L is the blade length. and These are normal stress and shear stress, respectively. For positive strain, For shear strain, The effective linear thermal expansion coefficient for functionally graded materials; For the temperature rise, , This represents the transient temperature field of the rotor blades. This refers to the initial temperature of the blade. y and z are the coordinates of the blade's length direction, thickness direction, and width direction, respectively; Potential energy generated by centrifugal stiffening Represented as: ; in, This represents the centrifugal force exerted on the cross-section at position x of the blade. This represents the lateral displacement of the blade. for First-order partial derivative with respect to coordinate x; Depend on and Constructing the total potential energy of the blade ,Right now .

8. The method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock according to claim 6, characterized in that, The work done by the unsteady aerodynamic loads on the blade throughout its entire motion time domain is as follows: ; In the formula, The work done by the unsteady aerodynamic loads over the entire time domain of the blade's motion.

9. The method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock according to claim 8, characterized in that, The differential equations of motion of the blade under unsteady aerodynamic loads and thermal shock are obtained based on the blade kinetic energy, total blade potential energy, and the work done by the unsteady aerodynamic loads throughout the blade's entire motion time domain. These equations include: The differential equations of motion for the rotor blades are derived from Hamilton's principle as follows: ; In the formula, The variational function is represented by t1 and t2, which are the start and end times of the analysis time interval, respectively. The blade kinetic energy, total blade potential energy, and the work done by the unsteady aerodynamic loads throughout the blade's entire motion time domain are substituted into the equation. Considering structural proportional damping, the differential equations of motion for the accelerated rotating aero-engine blades under unsteady aerodynamic loads and thermal shocks can be obtained: ; In the formula, , and These are the blade's generalized coordinate vector, generalized velocity vector, and generalized acceleration vector, respectively. , , , , , and These represent the blade mass matrix, gyroscope matrix, damping matrix, structural stiffness matrix, centrifugal stress stiffening matrix, rotational softening matrix, and stiffness matrix caused by angular acceleration, respectively. , and These are the generalized force vectors caused by rotational effects, thermal expansion effects, and aerodynamic loads, respectively.

10. The method for predicting vibration of accelerated rotating aero-engine blades under aerodynamic and thermal shock according to claim 9, characterized in that, The process of identifying resonance peaks from the time-domain response of bending displacement to determine the excitation frequency order, natural frequency order, and resonance type of the rotor blades includes: Identify resonance peaks and determine their occurrence times from the time-domain response of bending displacement; The response frequency at the resonance peak is extracted from the three-dimensional spectrum of the time-domain response to bending displacement. The response frequency at the resonance peak moment is matched with the instantaneous blade passage frequency and its harmonics, as well as the time-varying natural frequencies of each order, to determine the excitation frequency order, natural frequency order, and resonance type of the rotor blade.