CN117993125B - A macrodynamic model for the evolution of spalling fault in rolling element bearings of aeroengines - Google Patents

Abstract

The invention relates to the technical field of aeroengine fault diagnosis, in particular to a macroscopic dynamics model of aeroengine rolling bearing spalling fault evolution, which comprises the following steps: step S1, establishing a complete machine dynamics model of the aero-engine; step S2, deducing the variation of the contact deformation of the bearing caused by faults and the variation of the bearing force caused by the faults according to the geometric shape of the damage; s3, importing a rolling bearing fault macroscopic dynamics model to form an aero-engine complete machine vibration model for coupling the main bearing fault; s4, directly obtaining the vibration response of the whole machine caused by the fault of the rolling bearing by using a numerical integration method, and on the basis, simulating and analyzing the dynamic response rule in the bearing spalling fault evolution process; and S5, obtaining vibration acceleration of the vertical measuring points of the three-fulcrum main bearing seat and the intermediate case through numerical simulation so as to solve the problem that failure evolution behaviors cannot be considered in the prior art.

Description

A macrodynamic model for the evolution of spalling fault in rolling element bearings of aeroengines

Technical Field

The invention relates to the technical field of aircraft engine fault diagnosis, and in particular to a macroscopic dynamics model of the evolution of aircraft engine rolling bearing spalling fault.

Background technique

Rolling bearings, especially spindle bearings, are a very critical type of component in aircraft engines. They operate at high speeds and temperatures, experience drastic state changes, and are subject to complex time impact loads, making them extremely prone to failure. Once tiny early cracks appear, they will gradually expand into more obvious damage, and their failure will often cause the engine to shut down in the air or be returned to the factory prematurely. Therefore, establishing an accurate rolling bearing fault spalling dynamics model is of great significance for exploring accurate, efficient, and intelligent rolling bearing early fault diagnosis technology and remaining life prediction technology.

At present, classic single-point impact and multi-point impact fault dynamics models have been established for rolling bearings at home and abroad, and their typical signal characteristics have been analyzed. However, these classic methods do not consider the differences between faults of different sizes, and therefore cannot be used to simulate the evolution process of bearing faults, and cannot be directly used for aircraft engine main bearing fault evolution simulation.

In order to fully understand the fault evolution behavior of the main bearing in the complex large system of the engine, it is urgently necessary to develop a new rolling bearing macrodynamic model that can consider the evolution process of the fault from small to large, and on this basis, analyze the coupled response law of the entire engine under the excitation of the main bearing fault.

Summary of the invention

To this end, the present invention provides a macroscopic dynamic model of the evolution of spalling fault of an aircraft engine rolling bearing to overcome the problem that the fault evolution behavior cannot be considered in the prior art.

To achieve the above object, the present invention provides a macroscopic dynamic model of the evolution of spalling fault of an aero-engine rolling bearing, comprising the following steps:

Step S1, establishing a whole-machine dynamics model of the aircraft engine;

Step S2, the spalling failure of the rolling bearing is divided into two cases: raceway spalling and rolling element spalling damage. In the modeling of the inner and outer raceway spalling failure, the bearing damage is divided into two damage impact forms of "triangle" and "trapezoidal" according to the size of the damage. According to the geometric shape of the damage, the change in the bearing contact deformation caused by the fault and the change in the bearing force caused by the fault are derived;

Step S3, importing the established rolling bearing fault macro-dynamic model into the aircraft engine whole machine vibration model to form the aircraft engine whole machine vibration model coupled with the main bearing fault;

Step S4, using a numerical integration method to directly obtain the whole machine vibration response caused by the rolling bearing fault, so as to simulate and analyze the dynamic response law during the evolution of the bearing spalling fault;

Step S5, obtain the vibration acceleration of the three-point main bearing seat and the intermediate casing vertical measuring point through numerical simulation, extract the vibration effective value, kurtosis, and wavelet envelope characteristics from them, and then complete the verification of the established model.

Furthermore, the step S1 specifically includes:

Step S1-1, establishing a 5-DOF ball bearing dynamics model, and deriving bearing force and moment expressions under 5-DOF complex deformation;

Step S1-2, for the cylindrical roller bearing, using the "slice method", derive the force of the cylindrical roller bearing under the influence of the radial deformation of the bearing, the convexity of the cylindrical rotor, the bearing clearance and the angular deformation caused by the bearing tilt;

Step S1-3, combining the ball bearing model and the roller bearing model with the 6-DOF rotor and casing finite element beam model to establish an aircraft engine whole machine vibration model including rolling bearing modeling;

Step S1-4, using Method and an improved/> The differential equations are solved by combining the two methods, in which the method is used to solve the differential equations. The method is used to solve the finite element model of the rotor and casing which is easy to form a matrix, and the improved method is used to solve the finite element model of the rotor and casing which is easy to form a matrix. The method is used to solve the supporting connection components that do not need to form a matrix.

Furthermore, the improved The method is a new explicit integration method.

Furthermore, the step S2 specifically includes:

Step S2-1, assuming that the damage shape is a rectangular pit, whose cross section is the damage surface, assuming that LD is the diameter of the damage surface, a is the depth of the damage, _and rB is the radius of the _ball ;

Step S2-2, for the early stage of bearing surface damage, at this time, the damage area is small, the rolling element does not contact the bottom of the damage pit, and the impact formed at this time is a triangular impact. The transient displacement change generated by the rolling element when passing through the damage is:

;

Step S2-3: In order to obtain the impact position and impact magnitude of the rolling element on the raceway, it is necessary to consider the angular position of the damage on the bearing in different situations. Suppose the angular position of the jth rolling element is ,have , where/> is the rotation frequency of the cage; Z is the number of rolling elements;

Step S2-4, determining the gap amount of a key point on the circumference according to the damage position on the raceway, and then obtaining the gap amount change of the rolling body by interpolating between the key points according to the angular position of the rolling body on the circumference;

Step S2-5, for the evolution stage of bearing surface damage, at this time, the damage area continues to increase, and when the condition is met When the rolling element contacts the bottom of the damage pit, the transient displacement change process is in the form of a trapezoidal impact.

In the case of a trapezoidal impact, the rolling element will contact the bottom of the damage pit, and the depth of the damage pit is equal to the transient displacement change δ generated by the rolling element:

;

Step S2-6, considering the angular position of the damage on the bearing in different situations to obtain the impact position and impact amount of the rolling element on the raceway, assuming that the angular position of the jth rolling element is , there is/> , where/> is the rotation frequency of the cage; Z is the number of rolling elements; t is the cumulative time from the beginning of calculation using this formula;

Step S2-7, determining the clearance amount of a key point on the circumference according to the damage position on the raceway, and then obtaining the clearance amount change of the rolling body by interpolating between the key points according to the angular position of the rolling body on the circumference.

Furthermore, the step S4 specifically includes:

Step S4-1, using a numerical integration method to directly obtain the vibration response of the entire machine caused by the rolling bearing fault, and on this basis, simulate and analyze the dynamic response law during the evolution of the bearing spalling fault.

Furthermore, the step S5 specifically includes:

Step S5-1, the signal analysis method adopted is: using db8 wavelet as the basis to perform 5-layer wavelet decomposition to obtain 6 frequency band signals, namely: _Wd1 , _Wd2 , _Wd3 , _Wd4 , _Wd5 , _Wa5 ; assuming the signal sampling frequency is fs , the energy of each frequency band is: fs _/ 4 _- fs /2, fs _/ 8- _fs /4, fs /16 _- fs _/ 8, fs /32- fs /16 _, fs/64-fs _/ 32 _{, 0 -} fs / ₆₄ . /64; then, the Hilbert transform is used to obtain the envelope signal of each frequency band. In order to eliminate the interference of random signals, the autocorrelation denoising method is used to denoise the envelope signal of the frequency band decomposition signal; finally, FFT is used to obtain the wavelet envelope spectrum, and the bearing fault frequency characteristics are extracted from the spectrum; the effective value of the frequency band envelope signal is directly calculated to obtain the frequency band envelope energy characteristics FBEE1, FBEE2, FBEE3, FBEE4, FBEE5, and FBEE6.

Compared with the prior art, the beneficial effect of the present invention lies in that the present invention divides the spalling failure of the rolling bearing into two cases: raceway spalling and rolling element spalling damage. In the modeling of the inner and outer ring raceway spalling failure, the bearing damage is divided into two damage impact forms of "triangle" and "trapezoidal" according to the size of the damage. Among them, the early spalling failure mainly causes the "triangle" impact, while the middle and late failures cause the "trapezoidal" impact. According to the geometric shape of the damage, the change in the bearing contact deformation caused by the fault and the change in the bearing force caused by the fault are derived, and the vibration response of the whole machine caused by the rolling bearing fault is directly obtained by using the numerical integration method. On this basis, the dynamic response law of the bearing spalling fault evolution process is simulated and analyzed; the vibration acceleration of the three-point main bearing seat and the intermediate casing vertical measuring point is obtained through numerical simulation, and the vibration effective value, kurtosis, and wavelet envelope characteristics are extracted from them. The method of the present invention can accurately establish a rolling bearing fault evolution dynamic model, and can clearly separate the rolling bearing fault evolution state according to the frequency component through the signal analysis method, realize the identification of the fault evolution state and the early weak fault alarm of the rolling bearing, which is of great significance for the effective implementation of rolling bearing state monitoring, fault diagnosis and health management.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the overall structure of an engine according to the present invention;

FIG2 is a schematic diagram of a complete dynamics model of a twin-rotor aircraft engine of the present invention;

FIG3 is a flow chart of the complex rotor-support-casing coupling dynamics solution of the present invention;

FIG4 is a line graph of the actual measured values of the vibration velocity of the air intake casing measuring point of the present invention;

FIG5 is a modal calculation result of the whole machine at the critical speed of the present invention;

FIG6 is a critical speed test result of the intermediate casing measuring point of the present invention;

FIG7 is a simulation result of the whole machine model of the present invention;

FIG8 is an expanded view of bearing raceway damage according to the present invention;

FIG9 shows different positions of triangular damage on the raceway of the present invention;

FIG10 is a diagram showing a trapezoidal damage pattern caused by the rolling element passing through the raceway surface damage of the present invention;

FIG11 shows different positions of trapezoidal damage on the raceway of the present invention;

FIG12 is a fault feature extraction method of the present invention;

FIG13 is a graph showing a characteristic quantity variation trend (bearing seat measurement point) during the rolling bearing outer ring fault evolution process of the present invention;

FIG14 is a variation trend of characteristic quantities during the rolling bearing outer ring fault evolution process of the present invention (casing measurement point);

FIG15 is a graph showing a characteristic quantity variation trend during the rolling bearing inner ring fault evolution process (bearing seat measurement point) of the present invention;

FIG. 16 is a graph showing the variation trend of characteristic quantities during the evolution of the inner ring fault of the rolling bearing according to the present invention (casing measurement point).

Detailed ways

The following is a further description of the content of a novel macro-dynamic model of the spalling fault evolution of a rolling bearing of an aero-engine according to the present invention in conjunction with the accompanying drawings.

Step S1, establishing a whole-machine dynamics model of the aircraft engine;

Step S2, the spalling failure of the rolling bearing is divided into two cases: raceway spalling and rolling element spalling damage. In the modeling of the inner and outer ring raceway spalling failure, the bearing damage is divided into two damage impact forms of "triangle" and "trapezoidal" according to the size of the damage. The early spalling failure mainly causes the "triangle" impact, while the mid-to-late failure causes the "trapezoidal" impact. According to the geometric shape of the damage, the change in the bearing contact deformation caused by the fault and the change in the bearing force caused by the fault are derived;

Step S3, importing the established rolling bearing fault macro-dynamic model into the aircraft engine whole machine vibration model to form the aircraft engine whole machine vibration model coupled with the main bearing fault;

Step S4, using a numerical integration method to directly obtain the whole machine vibration response caused by the rolling bearing fault, and on this basis, simulate and analyze the dynamic response law during the evolution of the bearing spalling fault;

Step S5, obtain the vibration acceleration of the three-point main bearing seat and the intermediate casing vertical measuring point through numerical simulation, extract the vibration effective value, kurtosis, and wavelet envelope characteristics, and then verify the practicality of the established model.

Wherein, the step S1 specifically includes:

Step S1-1, establishing a 5-DOF ball bearing dynamics model, and deriving bearing force and moment expressions under 5-DOF complex deformation;

Step S1-2, for the cylindrical roller bearing, using the "slice method", derive the force of the cylindrical roller bearing under the influence of the radial deformation of the bearing, the convexity of the cylindrical rotor, the bearing clearance and the angular deformation caused by the bearing tilt;

Step S1-3, combining the ball bearing model and the roller bearing model with the 6-DOF rotor and casing finite element beam model to establish an aircraft engine whole machine vibration model including rolling bearing modeling;

Step S1-4, using Method and an improved/> The differential equations are solved by combining the two methods, in which the method is used to solve the differential equations. The method is used to solve the finite element model of the rotor and casing which is easy to form a matrix, and the improved method is used to solve the finite element model of the rotor and casing which is easy to form a matrix. The method is used to solve the supporting connection components that do not need to form a matrix.

Wherein, the step S1-1 specifically includes:

Step S1-1-1, the schematic diagram of the overall structure of a certain type of high thrust-to-weight ratio twin-rotor aircraft engine is shown in Figure 1, the calculation model of the rotor-support-casing-mounting node system is shown in Figure 2, the engine low-pressure rotor and high-pressure rotor and casing are simulated by beam units, the meanings of the dynamic connection symbols in Figure 2 are shown in Table 1, and the numbers 1, 3, 6-9, 11, 15-17, and 22 represent the node numbers used for modeling, among which the fan casing measurement point is set at node 3, the intermediate casing front measurement point is set at node 7, the intermediate casing rear measurement point is set at node 9, and the outer casing measurement point is set at node 16.

Table 1 Symbolic meanings of various connections defined in the kinetic model

Step S1-1-2, a certain type of twin-rotor aircraft engine has 5 pivots, of which the second and third pivots are angular contact ball bearings, and their structural parameters are shown in Table 2; the first, fourth and fifth pivots are cylindrical roller bearings, and their structural parameters are shown in Table 3; Table 4 is the typical simulation speed of the engine, and Table 5 is the axial force at the typical simulation speed of the engine; the three-pivot main bearing is taken as the research object to model and simulate the main bearing spalling fault, and Table 6 is the characteristic frequency of the engine three-pivot main bearing fault.

Table 2 Angular contact ball bearing models and parameters

Table 3 Cylindrical roller bearing models and parameters

Table 4 Engine simulation speed

Table 5 Axial force

Table 6 Characteristic frequency of three-point main bearing fault

Wherein, the step S1-4 specifically includes:

Step S1-4-1, using Method and an improved/> The differential equations are solved by combining the new explicit integration method (Zhai method) with the method of / > The method is used to solve the rotor and casing finite element models that are easy to form a matrix, and the Zhai method is used to solve the supporting connection parts that do not need to form a matrix; the characteristic of this method is that it only needs to assemble the dynamic matrix of a single rotor or casing component, and does not need to form a huge matrix of the entire system, and the solution efficiency is very high. The flow chart is shown in Figure 3.

In step S1-4-2, the response peak method is used to analyze the test data and simulate the whole machine model to identify the low-pressure excitation fan pitch modal critical speed within the low-pressure rotor operating speed range and the high-pressure excitation high-pressure compressor modal critical speed below the high-pressure rotor idling speed.

FIG4 is a line graph of the measured values of the vibration velocity of the intake casing measuring point, from which it can be seen that the engine has a critical speed between 6000-7000rpm; FIG5 is the finite element simulation result of the whole machine model, where (a) in FIG5 is the simulation result of the effective value of the vibration velocity of the intake casing measuring point. By comparing the variation law of the simulation value of the 1 times N1 component of the vibration velocity of the intake casing measuring point with the N1 speed, it can be clearly seen that the fan pitch under low-pressure excitation has two critical speeds between 6000-7000rpm, namely 6083rpm and 6775rpm; FIG5 (b) is the low-pressure excited fan and the high-pressure rotor in-phase pitch mode (N1=6083rpm), and FIG5 (c) is the low-pressure excited fan and the high-pressure rotor in-phase pitch mode (N1=6775rpm). It can be seen that the corresponding vibration mode at the critical speed of 6083rpm is the in-phase swing of the fan rotor and the high-pressure rotor; the corresponding vibration mode at the critical speed of 6775rpm is the anti-phase swing of the fan rotor and the high-pressure rotor.

Figure 6 is the measured value of the vibration velocity at the intermediate casing measuring point. It can be seen that the engine has a critical speed at 7618 rpm. (a) in Figure 7 is the simulation result of the effective value of the vibration velocity at the intermediate casing measuring point. By comparing the change law of the simulation value of the 1 times N2 component of the vibration velocity at the intermediate casing measuring point with the N2 speed, it can be seen that the high-pressure rotor pitch under high-pressure excitation has a critical speed at 7618 rpm. (b) in Figure 7 is the vibration shape of the whole machine at this critical speed.

Table 7 shows the comparison results of the calculated critical speed value of the finite element model of the whole machine and the measured value of the engine. Through the comparison, it can be seen that the calculated and measured errors of the pitch modal critical speed of the low-pressure excited fan within the working speed range are 6.24% and 4.44% respectively, and the calculated error of the high-pressure excited high-pressure rotor modal critical speed outside the working speed is 3.26%.

Table 7 Comparison between calculated and measured critical speeds of the whole beam unit model

The step S2 specifically includes:

Step S2-1, assuming that the damage shape is a rectangular pit, whose cross section is the damage surface, assuming that LD is the diameter of the damage surface, a is the depth of the damage, _and rB is the radius of the _ball ;

Step S2-2, for the early stage of bearing surface damage, at this time, the damage area is small, the rolling element does not contact the bottom of the damage pit, and the impact formed at this time is a triangular impact. The transient displacement change generated by the rolling element when passing through the damage is:

;

Step S2-3: In order to obtain the impact position and impact magnitude of the rolling element on the raceway, it is necessary to consider the angular position of the damage on the bearing in different situations. Suppose the angular position of the jth rolling element is ,have , where/> is the rotation frequency of the cage; Z is the number of rolling elements;

Step S2-4, determining the gap amount of a key point on the circumference according to the damage position on the raceway, and then obtaining the gap amount change of the rolling body by interpolating between the key points according to the angular position of the rolling body on the circumference;

Step S2-5, for the evolution stage of bearing surface damage, at this time, the damage area continues to increase, and when the condition is met When the rolling element contacts the bottom of the damage pit, the transient displacement change process is in the form of a trapezoidal impact.

In the case of a trapezoidal impact, the rolling element will contact the bottom of the damage pit, and the depth of the damage pit is equal to the transient displacement change δ generated by the rolling element:

;

Step S2-6: In order to obtain the impact position and impact magnitude of the rolling element on the raceway, it is necessary to consider the angular position of the damage on the bearing in different situations. Suppose the angular position of the jth rolling element is ,have , where/> is the rotation frequency of the cage; Z is the number of rolling elements; t is the cumulative time from the beginning of calculation using this formula;

Step S2-7, determining the clearance amount of a key point on the circumference according to the damage position on the raceway, and then obtaining the clearance amount change of the rolling body by interpolating between the key points according to the angular position of the rolling body on the circumference.

Wherein, the step S2-1 specifically includes:

Step S2-1-1, FIG8 is an expanded view of the bearing raceway damage; in FIG8, assume that the damage is a pit, the shape of which is a rectangular pit, and its cross section is the damage surface, assume that LD is _the diameter of the damage surface, a is the depth of the damage, _and rB is the radius of the ball.

Wherein, the step S2-2 specifically includes:

Step S2-2-1, for the early stage of bearing surface damage, at this time, the damage area is small, the rolling element does not contact the bottom of the damage pit, and the impact formed at this time is a triangular impact; when the condition is met When the rolling element rolls over the rolling surface damage, the center of mass will experience changes in A, B, C, D, and E, and the change in damage displacement is shown in Table 8. As can be seen from Figure 8, the transient change process of displacement at this time can be approximately considered as a triangular impact form.

Table 8 Transient changes in the process of mass center rolling over bearing surface damage

Wherein, the step S2-4 specifically includes:

The gap amount of a key point on the circumference is determined according to the damage position on the raceway, and then the gap amount change of the rolling body is obtained by interpolation between the key points according to the angular position of the rolling body on _the circumference; Figure 9 shows the different positions of the triangular damage on the raceway; where θ1 is _the angular position of the damage center point, which needs to be converted into an angle value between 0 and 2p; θ2 is half of the angle of the damage.

If the damage is on the outer ring,

,

If the damage is in the inner ring,

.

Table 9 shows the gap variation values of key points on the circumference in different situations, where the angles of the five key points on a circle from 0 to 2p increase sequentially.

Table 9 Clearance variation values of key points on the circumference of raceway triangle damage in different situations

Wherein, the step S2-5 specifically includes:

Step S2-5-1, for the evolution stage of bearing surface damage, at this time, the damage area continues to increase, and when the condition is met When the rolling element contacts the bottom of the damage pit, the transient displacement change process can be approximately considered as a trapezoidal impact form.

In the case of a trapezoidal impact, the rolling element will contact the bottom of the damage pit, and the transient displacement change of the rolling element is the depth of the damage pit. In Figure 10, the distance between point C1 and point C2 is:

.

Table 10 Transient changes in the process of mass center rolling over bearing surface damage

Wherein, the step S2-7 specifically includes:

It is necessary to determine the clearance of a key point on the circumference according to the damage position on the raceway, and then obtain the clearance change of the rolling body by interpolating between the key points according to the angular position of the rolling body on the circumference. Figure 11 shows different positions of trapezoidal damage on the raceway, where θ ₁ is the angular position of the damage center point, which needs to be converted into an angle value between 0 and 2p; θ ₂ is half of the top angle of the trapezoidal damage; θ ₃ is half of the hypotenuse angle of the trapezoidal damage;

(1) If the damage is on the outer ring,

,

;

(2) If the damage is in the inner ring,

,

.

Table 11 shows the gap variation values of key points on the circumference in different situations; among them, the angles of the 6 key points on a circle from 0 to 2p increase sequentially.

Table 11 Clearance variation values of key points on the circumference of raceway trapezoidal damage in different situations

The step S5 specifically includes:

Step S5-1, the signal analysis method adopted is mainly: using db8 wavelet as the basis to perform 5-layer wavelet decomposition to obtain 6 frequency band signals, namely: _Wd1 , _Wd2 , _Wd3 , _Wd4 , _Wd5 , _Wa5 ; assuming the signal sampling frequency is fs , the energy of each frequency band is: fs _/ 4 _- fs /2, fs /8- _fs /4, fs _/ 16 _- fs _/ 8, fs /32- fs /16, _fs /64 _- fs /32 _, 0 _- fs _/ 64 _. /64; Then, the envelope signal of each frequency band is obtained by Hilbert transform. In order to eliminate the interference of random signals, the envelope signal of the frequency band decomposition signal is denoised by the autocorrelation denoising method; Finally, on the one hand, the wavelet envelope spectrum is obtained by FFT, and the bearing fault frequency characteristics are extracted from the spectrum; on the other hand, the effective value of the frequency band envelope signal is directly calculated to obtain the frequency band envelope energy characteristics FBEE1, FBEE2, FBEE3, FBEE4, FBEE5, and FBEE6;

Step S5-2, simulation analysis of the evolution of three-point faults of aircraft engines; the results show that as the damage size increases, the effective value of the bearing seat measurement point gradually increases, and reaches the maximum value when the damage reaches 2mm. As the damage further increases, the effective value decreases instead, while the effective value of the casing measurement point has no increasing trend during the entire damage evolution process, indicating that the vibration effective value of the casing measurement point cannot be used as a monitoring value for the evolution of bearing faults; as the size of the spalling damage increases further, the dimensionless value of the fault frequency increases rapidly, and when the damage size reaches 0.2mm, the characteristic value reaches the maximum, and then, as the damage size increases further, the value remains basically unchanged, but there is a certain degree of fluctuation; as the size of the spalling damage increases further, the envelope energy characteristic values of different frequency bands all show different degrees of upward trends, indicating that it has a strong monitoring capability for the evolution of bearing faults.

The step S5-1 specifically includes:

Step S5-1-1: In the wavelet envelope spectrum, there are characteristic spectrum peaks near the outer race fault characteristic frequency f _o and its various order multiples. is the wavelet envelope spectrum value of the lth layer, l = 1, 2, 3, 4, 5, 6; assuming that the number of spectral lines of the envelope spectrum in the frequency range of 10 Hz—2 f _{0 is} Ne , then the average value of the envelope spectrum is:

;

Step S5-1-2, find the maximum value of the spectrum line at the fault characteristic frequency in the envelope spectrum, and set its characteristic frequency tolerance range to , the envelope spectrum interval is/> , then the number of frequency points within the tolerance range is:/> , the maximum fault frequency in the lth layer wavelet envelope spectrum is,

;

Step S5-1-3, construct a dimensionless feature quantity:

;

Step S5-1-4, calculate the dimensionless eigenvalues of the detail signals of each layer After that, it is still necessary to compare the eigenvalues of each detail signal and take its maximum value as the final eigenvalue, that is,

.

In order to study the changing trend of the monitoring characteristic quantity of rolling bearings during the evolution process, the size of the bearing outer ring spalling fault is set to expand continuously with time, as shown in Table 12, and the speed is N1=8880rpm, N2=14675rpm; the vibration acceleration of the bearing seat of the main bearing of the fulcrum 3 and the vertical measuring point of the intermediate casing are obtained through numerical simulation, and the effective value, kurtosis, and wavelet envelope characteristics of the vibration are extracted from them, as shown in Figure 13 and Figure 14 respectively; It can be seen from the figure that: 1) With the increase of damage size; For the bearing seat measuring point, the effective value gradually increases, and reaches the maximum value when the damage reaches 2mm. As the damage increases further, the effective value decreases instead; For the casing measuring point, there is no increasing trend in the effective value during the entire damage evolution process, indicating that the vibration effective value of the casing measuring point cannot be used as the shaft The monitoring value of the bearing fault evolution is 1) the outer ring and the casing measurement point can; 2) with the increase of damage size, the kurtosis value gradually increases, but is less than 3; 3) with the further increase of the outer ring spalling damage size, the dimensionless value of the outer ring fault frequency increases rapidly. When the damage size reaches 0.2mm, the characteristic value reaches the maximum. Then, with the further increase of the damage size, the value remains basically unchanged, but there is a certain degree of fluctuation; since there is no fault in the inner ring and rolling element, its value has been very small; 4) with the further increase of the outer ring spalling damage size, the envelope energy characteristic values of different frequency bands all show an upward trend to varying degrees, among which FBEE1, FBEE2, FBEE4, and FBEE5 are the most obvious; it shows that the frequency band envelope energy characteristics have a strong monitoring ability for the evolution of bearing faults.

Table 12 Evolution of bearing damage

In order to study the changing trend of the monitoring characteristic quantity of the rolling bearing during the evolution process, the size of the bearing outer ring spalling fault is set to expand continuously with time, as shown in Table 12, and the speed is N1=8880rpm, N2=14675rpm; the vibration acceleration of the three-point main bearing bearing seat and the intermediate casing vertical measuring point is obtained through numerical simulation, and the vibration effective value, kurtosis, and the envelope energy characteristics of each detail signal obtained after wavelet decomposition and reconstruction are extracted from them, as shown in Figures 15 and 16 respectively; It can be seen from the figure that: 1) With the increase of damage size; For the bearing seat measuring point, the effective value gradually increases, and reaches the maximum value when the damage reaches 2mm. As the damage increases further, the effective value decreases instead; For the casing measuring point, there is no increasing trend in the effective value during the entire damage evolution process, indicating that the vibration of the casing measuring point has The effective value cannot be used as a monitoring value for the evolution of bearing faults, while the casing measurement point can; 2) With the increase of damage size, the kurtosis value gradually increases, but is less than 3; 3) With the further increase of the inner ring spalling damage size, the dimensionless value of the inner ring fault frequency increases rapidly. When the damage size reaches 0.2 mm, the characteristic value reaches the maximum. Then, with the further increase of the damage size, the value remains basically unchanged, but there is a certain degree of fluctuation; since there is no fault in the outer ring and rolling element, its value has been very small; 4) With the further increase of the inner ring spalling damage size, the envelope energy characteristic values of different frequency bands all show an upward trend to varying degrees, among which FBEE1, FBEE2, FBEE4, and FBEE5 are the most obvious; indicating that the frequency band envelope energy characteristics have a strong monitoring capability for the evolution of bearing faults.

So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims (4)

1. A method for establishing a macroscopic dynamic model of the evolution of spalling failure of an aero-engine rolling bearing, characterized in that it comprises the following steps: Step S1, establishing a whole-machine dynamics model of the aircraft engine; Step S2, the spalling failure of the rolling bearing is divided into two cases: raceway spalling and rolling element spalling damage. In the modeling of the inner and outer raceway spalling failure, the bearing damage is divided into two damage impact forms of "triangle" and "trapezoidal" according to the size of the damage. According to the geometric shape of the damage, the change in the bearing contact deformation caused by the fault and the change in the bearing force caused by the fault are derived; Step S3, importing the established rolling bearing fault macro-dynamic model into the aircraft engine whole machine vibration model to form the aircraft engine whole machine vibration model coupled with the main bearing fault; Step S4, using a numerical integration method to directly obtain the whole machine vibration response caused by the rolling bearing fault, so as to simulate and analyze the dynamic response law during the evolution of the bearing spalling fault; Step S5, obtaining the vibration acceleration of the three-point main bearing seat and the vertical measuring point of the intermediate casing through numerical simulation, extracting the vibration effective value, kurtosis, and wavelet envelope characteristics from them, and then completing the verification of the established model; The step S1 specifically includes: Step S1-1, establishing a 5-DOF ball bearing dynamics model, and deriving bearing force and moment expressions under 5-DOF complex deformation; Step S1-2, for the cylindrical roller bearing, using the "slice method", derive the force of the cylindrical roller bearing under the influence of the radial deformation of the bearing, the convexity of the cylindrical rotor, the bearing clearance and the angular deformation caused by the bearing tilt; Step S1-3, combining the ball bearing model and the roller bearing model with the 6-DOF rotor and casing finite element beam model to establish an aircraft engine whole machine vibration model including rolling bearing modeling; Step S1-4, using Method and an improved/> The differential equations are solved by combining the two methods, in which the method is used to solve the differential equations. The method is used to solve the finite element model of the rotor and casing which is easy to form a matrix, and the improved method is used to solve the finite element model of the rotor and casing which is easy to form a matrix. The method is used to solve the supporting connection parts that do not need to form a matrix; The step S2 specifically includes: Step S2-1, assuming that the damage shape is a rectangular pit, whose cross section is the damage surface, assuming that LD is the diameter of the damage surface, a is the depth of the damage, _and rB is the radius of the _ball ; Step S2-2, for the early stage of bearing surface damage, at this time, the damage area is small, the rolling element does not contact the bottom of the damage pit, and the impact formed at this time is a triangular impact. The transient displacement change generated by the rolling element when passing through the damage is: , Step S2-3: In order to obtain the impact position and impact magnitude of the rolling element on the raceway, it is necessary to consider the angular position of the damage on the bearing in different situations. Suppose the angular position of the jth rolling element is ,have , where/> is the rotation frequency of the cage; Z is the number of rolling elements; Step S2-4, determining the gap amount of a key point on the circumference according to the damage position on the raceway, and then obtaining the gap amount change of the rolling body by interpolating between the key points according to the angular position of the rolling body on the circumference; Step S2-5, for the evolution stage of bearing surface damage, at this time, the damage area continues to increase, and when the condition is met When the rolling element contacts the bottom of the damage pit, the transient displacement change process is in the form of a trapezoidal impact. In the case of a trapezoidal impact, the rolling element will contact the bottom of the damage pit, and the depth of the damage pit is equal to the transient displacement change δ generated by the rolling element: , Step S2-6, considering the angular position of the damage on the bearing in different situations to obtain the impact position and impact amount of the rolling element on the raceway, assuming that the angular position of the jth rolling element is , there is/> , where/> is the rotation frequency of the cage; Z is the number of rolling elements; t is the cumulative time from the beginning of calculation using this formula; Step S2-7, determining the clearance amount of a key point on the circumference according to the damage position on the raceway, and then obtaining the clearance amount change of the rolling body by interpolating between the key points according to the angular position of the rolling body on the circumference. 2. The method for establishing a macroscopic dynamic model of the evolution of spalling failure of an aeroengine rolling bearing according to claim 1, characterized in that the improved The method is a new explicit integration method. 3. The method for establishing a macroscopic dynamic model of the evolution of spalling fault of an aero-engine rolling bearing according to claim 1, characterized in that step S4 specifically comprises: Step S4-1, using a numerical integration method to directly obtain the vibration response of the entire machine caused by the rolling bearing fault, and on this basis, simulate and analyze the dynamic response law during the evolution of the bearing spalling fault. 4. The method for establishing a macroscopic dynamic model of the evolution of spalling failure of an aircraft engine rolling bearing according to claim 1, characterized in that the step S5 specifically comprises: Step S5-1, the signal analysis method adopted is: using db8 wavelet as the basis to perform 5-layer wavelet decomposition to obtain 6 frequency band signals, namely: _Wd1 , _Wd2 , _Wd3 , _Wd4 , _Wd5 , _Wa5 ; assuming the signal sampling frequency is fs , the energy of each frequency band is: fs _/ 4 _- fs /2, fs _/ 8- _fs /4, fs /16 _- fs _/ 8, fs /32- fs /16 _, fs/64-fs _/ 32 _{, 0 -} fs / ₆₄ . /64; then, the Hilbert transform is used to obtain the envelope signal of each frequency band. In order to eliminate the interference of random signals, the autocorrelation denoising method is used to denoise the envelope signal of the frequency band decomposition signal; finally, FFT is used to obtain the wavelet envelope spectrum, and the bearing fault frequency characteristics are extracted from the spectrum; the effective value of the frequency band envelope signal is directly calculated to obtain the frequency band envelope energy characteristics FBEE1, FBEE2, FBEE3, FBEE4, FBEE5, and FBEE6.