Bearing defect quantification method based on energy mapping and physics modeling

Abstract

The invention relates to a bearing defect quantitative analysis technology, in particular to a bearing defect quantitative method based on energy mapping and physics modeling, which comprises the following steps of: acquiring vibration acceleration signals of a peeled rolling bearing at multiple positions in real time by using a plurality of vibration acceleration sensors; aiming at the collected acceleration signal, solving the time-frequency distribution of the vibration signal through a smooth pseudo Wigner-Ville distribution method; acquiring a boundary curve of energy peak value change of time-frequency distribution based on a marginal distribution method, and extracting an energy peak value and corresponding change time information from the boundary curve; establishing a rolling body-defect contact position model, and establishing a quantitative estimation model of the peeling area of the movable bearing based on the model; substituting the extracted time information into a quantitative estimation model of the peeling area of the movable bearing, and calculating to obtain the peeling defect size of the rolling bearing; according to the method, the defects of inaccurate estimation result and large error caused by neglecting the rolling body and peeling contact process in the traditional modeling method are effectively avoided.

Description

Technical Field

[0001] The present invention belongs to the technical field of bearing defect quantitative analysis, involving the technical fields of kinematics, dynamics, and physics, and particularly relates to a bearing defect quantitative method based on energy mapping and physics modeling. Background Art

[0002] Rolling bearings are known as the "joints of industry" and are widely used in important equipment such as aerospace, high-speed trains, and wind turbine generators. With the continuous development and progress of the industrialization process, higher requirements are put forward for the running accuracy, operation stability, and working reliability of rolling bearings. In actual industrial applications, rolling bearings are often subjected to cyclic stress, and fatigue peeling of materials is likely to occur on the raceway surface, forming spalling defects. Quantitative sizing of the raceway spalling can provide an objective quantitative index for predicting the remaining service life of rolling bearings, and has important practical engineering significance for bearing spare parts / repair plans and the condition-based maintenance of mechanical equipment.

[0003] Existing rolling bearing spalling quantitative techniques or methods focus on extracting the time interval of the double-impact characteristics caused by spalling defects in the vibration signals of rolling bearings, so as to achieve quantitative estimation of the spalling size. Through investigation, it is found that: (1) Existing methods mostly rely on experience to artificially select the time information associated with the rolling element entering and exiting the spalling area in the time-domain vibration signal, with low method efficiency and poor stability; (2) For the impact signals excited in the initial stage of the raceway spalling of rolling bearings, their characteristics are relatively weak and the signal-to-noise ratio is low. Existing methods are difficult to effectively and accurately extract the time information corresponding to the rolling element entering and impacting the spalling area, resulting in inaccurate defect quantitative estimation results; (3) Existing methods ignore the mapping mechanism between the mechanical parameter evolution mechanism caused by the contact of the rolling element with the front and rear edges of the spalling area and the energy distribution of the vibration impact signals excited by the contact. This mapping relationship directly and truly reflects the positional relationship between the rolling element and the spalling contact and the contact and extrusion degree. This positional relationship is directly related to the size of the spalling area.

[0004] The geometric contact relationship formed in each link of the contact between the rolling element and the spalling has an inestimable impact on the accuracy of the final spalling quantitative estimation result; in addition, the contact process between the rolling element and the spalling area is always accompanied by the strong and weak conversion of kinetic energy and potential energy, which provides a new way and approach for the quantitative evaluation algorithm of the spalling area, and is expected to provide interpretability for this method from the perspective of energy transformation. However, there is still no method for electrically detecting bearing defects from the perspective of energy change in the existing technology. Summary of the Invention

[0005] In order to effectively quantitatively estimate the raceway spalling of rolling bearings, the present invention proposes a bearing defect quantitative method based on energy mapping and physics modeling, specifically including the following steps:

[0006] Use multiple vibration acceleration sensors to collect vibration acceleration signals of spalling rolling bearings in real time at multiple positions;

[0007] For the collected acceleration signals, obtain the time-frequency distribution of the vibration signals through the smoothed pseudo Wigner–Ville distribution method;

[0008] Based on the edge distribution method, obtain the boundary curve of the energy peak change of the time-frequency distribution, and extract the energy peak and the corresponding change time information from it;

[0009] Establish a rolling element-defect contact position model, and based on this model, establish a quantitative estimation model for the spalling area of dynamic bearings;

[0010] Substitute the extracted time information into the quantitative estimation model for the spalling area of dynamic bearings to calculate the size of the spalling defect of the rolling bearing.

[0011] The advantages of the present invention compared with the prior art are as follows:

[0012] 1. The present invention constructs an energy change-time mapping relationship by taking the extraction of key time information in multiple stages of rolling element defect contact as the starting point. The present invention overcomes the shortcomings of low quantization accuracy of defect size and large result variance caused by empirical selection of time information in traditional methods. It is applicable to the effective extraction of the time information corresponding to the key angular positions of the rolling element in contact with the spalling under different working conditions of the rolling bearing, and solves the problems such as the complex process of extracting the time information in multiple stages of the rolling element in contact with the spalling in traditional methods.

[0013] 2. The present invention is based on the geometric and physical collaborative equivalent modeling of the multi-process of rolling element-spalling contact and the defect quantization model. The present invention overcomes the problems of insufficient calculation accuracy and being affected by the rotational speed caused by establishing the defect size quantization model by simplifying the assumption of the path of the rolling element passing through the spalling area in traditional methods, and effectively avoids the shortcomings of inaccurate estimation results and large errors brought by ignoring the rolling element-spalling contact process in traditional modeling methods. Description of the Drawings

[0014] Figure 1 Schematic diagram of the contact relationship between the rolling element and the raceway spalling in the present invention;

[0015] Figure 2 Flow chart of quantitative estimation of raceway spalling in the present invention;

[0016] Figure 3 Vibration acceleration signal diagram in the present invention;

[0017] Figure 4 Vibration shock signal and its instantaneous energy diagram in the present invention;

[0018] Figure 5Instantaneous energy curve diagram of the vibration acceleration signal and its smoothed pseudo-Wigner–Ville time-frequency distribution in the present invention;

[0019] Figure 6 Box plot of the quantitative estimation results of the spalling zone width at different rotational speeds in the present invention. Specific implementation manners

[0020] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] The present invention proposes a quantitative method for bearing defects based on energy mapping and physical modeling, which specifically includes the following steps:

[0022] Use multiple vibration acceleration sensors to perform multi-position real-time acquisition of the vibration acceleration signal of a spalling rolling bearing;

[0023] For the acquired acceleration signal, obtain the time-frequency distribution of the vibration signal through the smoothed pseudo-Wigner–Ville distribution method;

[0024] Based on the edge distribution method, obtain the boundary curve of the energy peak change of the time-frequency distribution, and extract the energy peak and the corresponding change time information therefrom;

[0025] Establish a rolling element-defect contact position model, and based on this model, establish a quantitative estimation model for the spalling zone of a dynamic bearing;

[0026] Substitute the extracted time information into the quantitative estimation model for the spalling zone of a dynamic bearing to calculate the size of the spalling defect of the rolling bearing.

[0027] Next, this embodiment will be combined with Figures 1 to 6 to illustrate the solution of the present invention.

[0028] In this embodiment, a piezoelectric vibration acceleration sensor is installed on the bearing housing where the test bearing is located. With the help of a 4-channel high-sampling-rate data acquisition card, it is connected to 3 vibration acceleration sensors and 1 eddy current sensor, and the other end of the data acquisition card is connected to a computer. Through the operation of the acquisition program, the vibration acceleration signals of the spalling rolling bearing in the horizontal and vertical directions are acquired in real time to provide accurate, complete and effective source signals for subsequent analysis.

[0029] The autoregressive model is used to perform impact feature enhancement processing on the vibration acceleration signal collected in the previous step. Based on a large number of literature researches, the maximum order P of the autoregressive model is set to 200, and the collected vibration acceleration signal A(t) is filtered by the autoregressive model based on the maximum kurtosis to obtain the vibration acceleration signal R(t) with enhanced impact features after autoregressive filtering.

[0030] The mathematical model of autoregressive model filtering is as follows:

[0031]

[0032] In the formula, A(t - n) is the value of the original vibration signal A(t) at its time position t - n, t represents the time index position of the signal itself, P is the maximum set order of the autoregressive model, a n (n = 1, 2, 3, …, 200) is the autoregressive coefficient, and R(t) is the residual signal after autoregressive model filtering, which mainly contains the noise signal and the impact feature signal of the faulty rolling bearing.

[0033] By calculating the kurtosis value K p of the output R p (t) of the autoregressive model corresponding to each order P (P = 1, 2, 3, …, 200), that is:

[0034]

[0035] In the formula, R p (t) is the residual signal after autoregressive filtering, and the subscript represents the index value of the vibration signal under different filtering order p conditions; R m is the mean value of the signal R p (t), L is the signal length, and ρ is the signal standard deviation.

[0036] By calculating the kurtosis value K P of the filtered signal corresponding to different filtering orders P, comparing and obtaining the maximum kurtosis value K max , and using the order P corresponding to K max as the optimal filtering order of the autoregressive model to filter the signal, the optimal vibration acceleration signal R(t) with enhanced impact features is obtained. The optimal filtering result diagram is as shown in Figure 3 shown. This diagram records the partial time period signal (0.022s - 0.174s) of the signal of the collected vibration acceleration signal after autoregressive filtering based on the maximum kurtosis. This diagram shows that the rolling of the rolling element over the spall will excite impact vibration characteristics, and this impact (vibration) signal exhibits periodic characteristics with the fault period as the time interval, as shown by the amplitude of the sharp vibration signal in the figure.

[0037] The time-frequency distribution can well describe the energy density distribution of non-stationary signals and can give a method for estimating the energy density distribution of signals. The present invention adopts a smoothed pseudo Wigner–Ville time-frequency distribution analysis method that can eliminate cross terms in vibration signals and achieve the positive energy distribution characteristic, and performs time-frequency analysis on the autoregressive filtered vibration acceleration signal with the maximum kurtosis obtained after the previous step of filtering. The theoretical calculation formula of the smoothed pseudo Wigner–Ville time-frequency distribution analysis method is:

[0038]

[0039] In the formula, R(t + 0.5τ) represents the analytic form of the vibration acceleration signal R(t), R* represents the complex conjugate of R(t), and τ represents the delay of time t; h(τ) is a time window function that can suppress the signal cross terms in the frequency domain of the time-frequency distribution; g(u - τ) is a frequency window function that can suppress the signal cross terms in the time domain of the time-frequency distribution. In the present invention, the standard Gauss window function is selected as the window function in the analysis method, and its expression is:

[0040]

[0041] where W(n) is the output signal of the window function, σ is the standard deviation of the input window function processing signal x, and μ is the variance of the input window function processing signal x. The time-frequency distribution diagram obtained by calculating the filtered vibration acceleration signal through the smoothed pseudo Wigner–Ville distribution is as shown in Figure 4 shown. The figure records the time-frequency analysis of a partial signal intercepted from the vibration impact characteristic signal shown in Figure 3 , indicating the frequency characteristics of the impact characteristics generated when the rolling element enters, impacts, and exits the defect, and the relationship between the change in energy density contained therein and time.

[0042] To accurately extract the time information corresponding to the change in energy peak caused by the whole process of rolling element-spall contact, using the edge distribution calculation method of the time-frequency distribution characteristics of the smoothed pseudo Wigner–Ville energy density, the instantaneous energy change curve of the energy density of the smoothed pseudo Wigner–Vill time-frequency distribution of the vibration acceleration signal is calculated, that is:

[0043]

[0044] In the formula, |E I (t)| 2Represents the instantaneous energy of the signal. By obtaining the characteristic curve of the change in instantaneous energy, the key time information corresponding to when the rolling element starts to contact the spalling area, starts to enter the spalling area, and impacts the spalling area at the trailing edge of the spalling area can be obtained. This provides crucial time input parameters for the next-step estimation model. Among them, the curve of the change in the instantaneous energy of the vibration shock signal is as Figure 5 shown. The figure records the curve of the change in the instantaneous energy of the filtered vibration shock signal and its corresponding smoothed pseudo-Wigner–Ville energy density distribution, indicating the energy change characteristics of the shock signal generated when the rolling element passes through the defect and its mapping relationship with time.

[0045] Next, in combination with Figure 1 the schematic diagram of the contact action between the rolling element and the outer-raceway spall shown, the process of establishing the rolling element-defect contact position model and the quantitative estimation model of the spalling area of the rolling bearing based on this model will be described.

[0046] As Figure 1 shown is the schematic diagram of the contact action between the rolling element and the outer-raceway spall. Among them, L is the leading-edge position of the defect area, P s is the center position of the rolling element in the outer-raceway without defects, P a1 is the center position of the rolling element at the moment before contacting the spalling area, P d is the center position of the rolling element at the leading edge of the spalling area, P i is the center position of the rolling element at the trailing edge of the spalling area. The connecting lines between each position and the outer-raceway are respectively OP s , OP a1 , OP d , OP i . The connecting line between the outer-raceway and the contact point T at the trailing edge of the spalling area is OT. Among them, the angle between OP s and OP a1 is ψ b , the angle between OP a1 and OP d is ψ d , the angle between OP d and OP i is ψ di , the angle between OP a1 and OT is ψ spall ; The distance from the contact point C at the leading edge of the spalling area to the center O of the outer-raceway can be expressed as:

[0047] OC = (0.5D p + 0.5D b )-(δ br + 0.5C d ) (6)

[0048] In the formula, Db Denotes the diameter of the rolling element, D p Denotes the pitch circle diameter of the rolling bearing, δ br Is the total contact deformation between the rolling element and the inner and outer rings of the bearing, C d Is the radial clearance of the rolling bearing.

[0049] In ΔOCL, due to the angle ψ b Is very small, according to the trigonometric relationship, the following equation relationship can be obtained:

[0050]

[0051] In the formula, A ou Denotes the length of the major semi - axis of the ellipse formed by the contact between the rolling element and the outer ring.

[0052] Let:

[0053] L OC = OC = 0.5(D p + D b ) - (δ br +0.5C d ) (8)

[0054] In ΔOLP d According to the cosine law, the angle ψ d Can be defined as the following relationship:

[0055] 2L OC D p cosψ d -(L OC ) 2 -(0.5D p ) 2 +(0.5D b ) 2 =0 (9)

[0056] Rearranging the above formula gives:

[0057] L OC D p -(D b -δ br -0.5D d )(δ br +0.5D d )-L OC D p cosψ d =0 (10)

[0058] Let G d =(D b -δ br -0.5C d )(δbr +0.5C d ), according to the trigonometric function relationship, we can get:

[0059]

[0060] Since D ball >>δ br , δ br <<1 and ψ d is very small, so the above formula can be further simplified to:

[0061]

[0062] Based on the above derivation, the starting disconnection time interval t sd is given by the following equation:

[0063] ψ b +ψ d = ω cage t sd (13)

[0064] The angular range of the rolling element from position P d to P i , ψ di is very small. Therefore, it can be assumed that the speed of the center of the rolling element from P d to P i is constant, and we can get:

[0065] ψ di =ω cage t di (14)

[0066] When the rolling element moves from position P d to P i , the distance moved by the rolling element in the vertical direction can be calculated by the following relationship:

[0067]

[0068] According to the above derivation and Figure 1 the geometric contact relationship shown, the following relational expression can be established:

[0069] [OTcosψ spall -(OP d +H d )cos(ψ d +ψ di )] 2 +[OTsinψ spall -(OP d +H d )sin(ψ d +ψdi )] 2 =(0.5D b ) 2 (16)

[0070] Wherein, OT and OP d can be expressed as 0.5(D p +D b )-(δ bi +δ bo ) and 0.5(D P +D b ).

[0071] Finally, the size of the spalling defect of the rolling bearing (i.e., the length along the circumferential direction of the outer ring) W spall can be expressed by the following formula:

[0072]

[0073] In this embodiment, the time information of the key angular positions of the mapped rolling elements extracted in the impact feature instantaneous energy distribution and edge extraction module is used. The time information includes the moment t c when the rolling element just contacts the front edge of the defect, the moment t e when the rolling element disengages from the contact with the front edge of the defect, and the moment t i when the rolling element impacts the rear edge of the defect. Through these three moment information, the time information t d from the angular position P i of the rolling element to P di can be calculated, expressed as t di =t e -t c . The time information t s from the angular position P d of the rolling element to P sd can also be calculated, expressed as t sd =t i -t c . Substitute the input parameters into the rolling bearing outer ring roll spalling quantitative estimation model derived and established in the spalling quantitative mathematical model construction module, and finally calculate and solve to obtain the rolling bearing spalling quantitative estimation size or width value. By applying the method proposed in the present invention and combining with the calculation flow chart shown in Figure 2 , the vibration impact signals of the rolling bearing with a spalling width of 0.93 mm on the outer ring are analyzed and calculated under 7 different rotational speed conditions, and the estimation results are shown in Figure 6 , Figure 6Recorded are the results of quantitative estimation of spalling for a rolling bearing with an outer raceway spalling of 0.93 mm under 7 different rotational speed conditions by using the method proposed in the present invention, indicating that the method proposed in the present invention can effectively estimate the width dimension of the spalling area quantitatively and has good robustness and stability.

[0074] As an alternative embodiment, as Figure 5 , in this figure, the abscissa of the vibration acceleration signal is time (unit: second s), the ordinate is vibration acceleration (unit: meter per square second m / s 2 ), the abscissa of the instantaneous energy curve is time (unit: second s), the ordinate is energy. In this embodiment, according to the instantaneous energy distribution of the impact characteristics, the moment t corresponding to the rolling element just contacting the front edge of the defect is determined c , the moment t corresponding to the rolling element separating from the contact with the front edge of the defect e and the moment t corresponding to the impact of the rolling element with the rear edge of the defect i These three moment information. First, the instantaneous energy curve of the rolling element is obtained according to the vibration acceleration signal of the rolling element. The moment when the instantaneous energy curve of the rolling element does not drop to 0 for the first time after the fluctuation and the instantaneous energy curve rises at the next moment is taken as the moment t corresponding to the rolling element just contacting the front edge of the defect c , the moment when the instantaneous energy curve first appears the maximum value of the instantaneous energy is taken as the moment t corresponding to the rolling element separating from the contact with the front edge of the defect e , and the moment when the subsequent maximum value of the instantaneous energy appears in the instantaneous energy curve and the maximum value of the instantaneous energy is greater than the maximum value of the instantaneous energy for the first time is taken as the moment t corresponding to the impact of the rolling element with the rear edge of the defect i .

[0075] The present invention also proposes a bearing defect quantitative system based on energy mapping and physical modeling for implementing a bearing defect quantitative method based on energy mapping and physical modeling, including:

[0076] A signal acquisition module, which uses a data acquisition card with signal conditioning function and a vibration acceleration sensor to collect the vibration acceleration signals of the spalling rolling bearing at multiple positions in real time to provide accurate and complete source signals for subsequent analysis;

[0077] A signal enhancement module, which filters the original signal by using an autoregressive model based on maximum kurtosis to enhance the characteristics of the impact signals corresponding to the spalling in the vibration signals, and finally enhances the signal-to-noise ratio to provide vibration acceleration signals with enhanced impact characteristics for subsequent processing;

[0078] The time information extraction module uses the smoothed pseudo Wigner–Ville distribution to obtain the time-frequency distribution of the vibration acceleration signal with enhanced impact features extracted in the previous step, and obtains the energy density change features of the impact signals excited during the whole process of rolling element-spall contact. Then, using the marginal distribution, a curve graph showing the instantaneous energy feature changes of the vibration impact signals corresponding to each stage of rolling element-spall contact is extracted, and the critical moment information corresponding to the energy peak is adaptively extracted from this instantaneous energy change curve graph based on the peak adaptive extraction method;

[0079] The bearing spall area quantitative estimation module, based on methods and theories such as physics, kinematics, and mechanics, combines the size parameters of the rolling bearing itself to establish a geometric contact relationship equation set formed by rolling element spall contact, and solves to obtain a rolling bearing spall quantitative estimation function model with the time information from the previous step as the independent variable. Using the time information corresponding to the peak of the instantaneous energy change of the vibration impact signal extracted above as the input parameter, it is substituted into the established mathematical model for quantitative estimation of rolling bearing raceway spall to achieve adaptive quantitative estimation of the spall area.

[0080] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A quantitative method for bearing defects based on energy mapping and physics modeling, characterized in that, Specifically, it includes the following steps: Use multiple vibration acceleration sensors to collect the vibration acceleration signals of the spalling rolling bearing in real time at multiple positions; For the collected acceleration signals, obtain the time-frequency distribution of the vibration signals by the smoothed pseudo Wigner–Ville distribution method; Based on the edge distribution method, obtain the boundary curve of the energy peak change of the time-frequency distribution, and extract the energy peak and the corresponding change time information from it; Establish a rolling element-defect contact position model, and based on this model, establish a quantitative estimation model for the spalling area of the dynamic bearing; Substitute the extracted time information into the quantitative estimation model for the spalling area of the dynamic bearing, and calculate the spalling defect size of the rolling bearing; 2. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 1, characterized in that, Filter the collected acceleration signals by using the autoregressive model based on the maximum kurtosis; 3. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 1 or 2, characterized in that Obtaining the time-frequency distribution of the vibration signals by the smoothed pseudo Wigner–Ville distribution method includes: Among them, SPWVD A (t, f) represents the time-frequency distribution function of the vibration signal, f represents the instantaneous frequency; h(τ) is the window function; R(t) is the vibration signal to be processed that changes with time t, R * (t) represents the complex conjugate of R(t), and R(t±τ) represents the signal obtained by shifting R(t) forward or backward by τ units at time t; g(u - τ) is the frequency window function, and u represents the independent variable.

4. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 1, characterized in that The instantaneous energy of the time-frequency distribution is expressed as: Among them, SPW A (t, f) represents the smoothed pseudo Wigner–Ville function, where t represents the time variable and f represents the frequency variable; f1 is the lower limit of the integration frequency; f2 is the upper limit of the integration frequency.

5. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 1, characterized in that When the rolling element starts to contact the spalling area position, starts to roll into the spalling area position, and contacts the rear edge of the spalling area, the contact mode between the rolling element and the defect changes. By obtaining the change curve of the instantaneous energy, the key time information corresponding to the position where the rolling element starts to contact the spalling area, starts to enter the spalling area, impacts the rear edge of the spalling area, and completely exits the spalling area can be obtained; 6. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 1, characterized in that, Based on the rolling element-defect contact position model, obtain the angle between the line connecting the contact point at the front edge of the spall zone and the outer raceway and the line connecting the contact point at the rear edge of the spall zone and the outer raceway, that is, the defect angle span ψ spall , and the rolling element-defect contact position model is expressed as: [OT cos ψ spall -(OP d +H d ) cos(ψ d + ψ di )] 2 + [OT sin ψ spall -(OP d +H d ) sin(ψ d + ψ di )] 2 =(0.5D b ) 2 OT = 0.5(D p + D b ) - (δ bi + δ bo ) OP d = 0.5(D P + D b ) Among them, OT is the connecting line distance between the center of the outer raceway and the position where the rolling element leaves the spalling area; OP d is the connecting line distance between the center of the outer raceway and the position where it just enters the spalling area; ψ spall is the defect angle span; ψ d is the angular span between the position of the rolling element immediately before entering the spalling area and the position reaching the leading edge of the spalling area; ψ di is the angular span between the position where the rolling element is just entering the spalling area and the position immediately before leaving the spalling area; D b is the diameter of the rolling element; D p is the pitch circle diameter of the rolling bearing; δ bi is the contact deformation between the rolling element and the inner race; δ bo is the contact deformation between the rolling element and the outer race; δ hd is the total contact deformation between the rolling element and the inner and outer races; t di is the time span for the rolling element to move from the leading edge position of the spalling area to the trailing edge position of the spalling area; V b is the linear velocity of the rolling element movement.

7. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 6, characterized in that The spalling defect size of the rolling bearing is expressed as:

8. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 6, characterized in that The angular span ψ of the rolling element moving from the front edge position of the spalling area to the rear edge position of the spalling area di is expressed as: ψdi = ω cage tdi Among them, ω cage represents the angular velocity of the cage.

9. A quantitative method for bearing defects based on energy mapping and physics modeling according to claim 6, characterized in that, The position of the rolling element at the moment before entering the spalling zone and the angular span ψ to the position reaching the leading edge of the spalling zone d Expressed as: L OC = 0.5(D p + D b ) - (δ br + 0.5C d ) G d = (D b - δ br - 0.5C d )(δ br + 0.5C d ) Among them, L OC is the distance from the contact point C at the front edge of the spalling area to the center O of the outer ring raceway; C d is an intermediate variable, expressed as G d =(D ball -δ br -0.5C d )(δ br +0.5C d ); δ br is the total contact deformation between the rolling element and the inner and outer rings of the bearing; C d is the radial clearance of the rolling bearing.

Images