Aerospace bearing cage rotational speed ultrasonic monitoring method

Abstract

The application relates to an aerospace bearing retainer rotating speed ultrasonic monitoring method and belongs to the technical field of bearing retainer rotating speed measurement. The method can solve the problem that the existing ultrasonic monitoring method for retainer rotating speed cannot accurately measure the rotating speed of a rolling bearing used in aerospace. The method comprises the following steps: obtaining a reflected pulse signal, performing Fourier transform to obtain ultrasonic normalized amplitude under a pulse center frequency, performing typical pit signal contraction transform to obtain normalized ultrasonic frequency domain amplitude after typical pit contraction transform, performing zero mean processing on the ultrasonic frequency domain amplitudes before and after the transform, obtaining power density frequency spectrums corresponding to the ultrasonic frequency domain amplitudes before and after the transform, determining a power density peak value after the transform according to a frequency corresponding to the power density peak value before the transform, calculating a peak value proportion ratio before and after the transform, obtaining a frequency corresponding to a maximum peak value proportion ratio before and after the transform as a rolling body passing frequency, and calculating the retainer rotating speed based on the rolling body passing frequency. The application is used for monitoring the rotating speed of an aerospace bearing retainer.

Description

Technical Field

[0001] This invention relates to an ultrasonic monitoring method for the rotational speed of aerospace bearing cages, belonging to the field of bearing cage rotational speed measurement technology. Background Technology

[0002] Bearing slippage monitoring is of core value in ensuring the reliability of rotating machinery. Slippage can cause the lubricating film between the rolling elements and raceways to rupture, leading to localized stress concentration and surface damage, accelerating fatigue spalling and vibration instability. Cage speed is a key indicator for assessing the transient slippage characteristics of bearings. By monitoring slippage characteristic parameters such as cage speed in real time, bearing operating performance can be evaluated in real time, bearing failures can be predicted, and thus provide key data support for improving the reliability, efficiency, and lifespan of mechanical systems, driving the transformation of equipment management from "post-failure maintenance" to "intelligent prevention."

[0003] Traditional cage speed monitoring primarily relies on fiber optic sensors. This method is susceptible to severe interference from oil mist, leading to distorted measurement results and difficulty in accurately determining the true slippage state of the bearing. In recent years, ultrasonic monitoring technology has attracted attention due to its high penetration and immunity to oil mist interference. Researchers have already used ultrasonic pulse time-domain amplitude and frequency-domain amplitude for cage speed monitoring. However, for aerospace rolling bearings operating at speeds exceeding 2 million dn (mm), the ultrasonic pulse transmitter and receiver hardware has limited transmission frequency, making it difficult for the ultrasonic pulse to scan every passing rolling element. This results in a reduced monitored rolling element frequency compared to the actual frequency, leading to significant errors in cage speed measurement. Summary of the Invention

[0004] To address the problem that existing ultrasonic monitoring methods for cage rotation speed cannot accurately measure the rotation speed of rolling bearings used in aerospace applications, this invention provides an ultrasonic monitoring method for the rotation speed of aerospace bearing cages.

[0005] An ultrasonic monitoring method for the rotational speed of an aerospace bearing cage according to the present invention includes:

[0006] An ultrasonic probe was used to acquire the time series of reflected pulse signals as the bearing rolling elements rolled, and Fourier transform was performed to obtain the time series of ultrasonic normalized amplitude at the pulse center frequency.

[0007] The normalized ultrasonic amplitude sequence is processed, and the typical crater signal over which the rolling body rolls is subjected to a contraction transformation to obtain the normalized ultrasonic frequency domain amplitude time sequence after the typical crater contraction transformation at the pulse center frequency.

[0008] The normalized ultrasonic amplitude time series and the normalized ultrasonic frequency domain amplitude time series after typical pit shrinkage transformation are respectively subjected to zero-mean processing to obtain the zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series.

[0009] The zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series are processed to obtain the normalized power density spectrum of the time series before transformation and the normalized power density spectrum of the time series after transformation. The power density peak value and corresponding frequency of the normalized power density spectrum of the time series before transformation are extracted, and the power density peak value of the normalized power density spectrum of the time series after transformation is determined by the frequency.

[0010] Calculate the percentage of the peak power density in the normalized power density spectrum of each time series before transformation, and the percentage of the peak power density in the normalized power density spectrum of each time series after transformation.

[0011] The calculation range of the rolling element passing frequency is determined based on the theoretical value of the rolling element passing frequency. The peak percentages before and after all transformations are filtered to obtain the filtered peak percentages before and after transformations. The peak percentages after transformations are divided by the corresponding points of the filtered peak percentages before transformations, and the frequency corresponding to the maximum value of the peak percentage is taken as the rolling element passing frequency.

[0012] The cage rotational speed is calculated based on the frequency of the rolling elements.

[0013] According to the ultrasonic monitoring method for the rotational speed of aerospace bearing cages of the present invention, the time series of reflected pulse signals for:

[0014] ,

[0015] In the formula For time, This is the first reflected pulse signal. This is the second reflected pulse signal. This is the third reflected pulse signal;

[0016] Time series of reflected pulse signals Perform a Fourier transform to obtain the normalized amplitude time series of ultrasound at the pulse center frequency. for:

[0017] ,

[0018] In the formula The pulse center frequency, Indicates Fourier transform, Indicates the frequency at the center of the pulse. The value to be taken below, This indicates the amplitude value.

[0019] According to the ultrasonic monitoring method for aerospace bearing cage speed of the present invention, the normalized ultrasonic frequency domain amplitude time series after transformation of typical dimple contraction at the pulse center frequency is expressed as follows: :

[0020] ,

[0021] In the formula The threshold for determining whether there are typical dents caused by a rolling element in the normalized amplitude time series of ultrasound.

[0022] According to the ultrasonic monitoring method for aerospace bearing cage speed of the present invention, the zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series are as follows:

[0023] ,

[0024] In the formula This is a zero-mean normalized ultrasound amplitude time series. This is the transformed zero-mean normalized ultrasonic frequency domain amplitude time series. for The mean, for The mean;

[0025] .

[0026] According to the ultrasonic monitoring method for aerospace bearing cage rotation speed of the present invention, the normalized power density spectrum of the time series before transformation is represented as follows: The normalized power density spectrum of the transformed time series is expressed as: Threshold PSD is determined by setting the peak power density. ref ,get:

[0027] ,

[0028] In the formula This is an array of power density peaks in the normalized power density spectrum of the time series before transformation. This is the power density peak array of the normalized power density spectrum of the transformed time series. For frequency The corresponding peak power density of the normalized power density spectrum of the time series before transformation, where k is the number of peaks. for The corresponding power density peak value of the normalized power density spectrum of the transformed time series.

[0029] According to the ultrasonic monitoring method for aerospace bearing cage rotation speed of the present invention, the method for calculating the percentage of peak power density in the normalized power density spectrum of the time series before transformation and the percentage of peak power density in the normalized power density spectrum of the time series after transformation are as follows:

[0030] ,

[0031] In the formula This is an array of peak percentages before the transformation. This is an array of peak percentages after transformation.

[0032] According to the ultrasonic monitoring method for the rotational speed of aerospace bearing cages of the present invention, the theoretical value of the rolling element passing frequency is determined based on the measured inner ring rotational speed of the bearing. 10% of the theoretical value is taken as the lower limit of the calculation range of the rolling element passing frequency, and 120% of the theoretical value is taken as the upper limit of the calculation range of the rolling element passing frequency.

[0033] According to the ultrasonic monitoring method for aerospace bearing cage speed of the present invention, the method for calculating the rolling element passing frequency is as follows:

[0034] ,

[0035] In the formula To increase the maximum value of the obtained peak percentage, For the frequency of the rolling element, This is a filtered array of transformed peak percentages. This is a filtered array of peak percentages before transformation.

[0036] According to the ultrasonic monitoring method for aerospace bearing cage rotation speed of the present invention, the calculation method for cage rotation speed is as follows:

[0037] ,

[0038] In the formula To maintain the rack speed, This represents the number of rolling elements.

[0039] According to the ultrasonic monitoring method for aerospace bearing cage rotation speed of the present invention, the normalized power density spectrum of the time series before transformation is... and the normalized power density spectrum of the transformed time series The method to obtain it is as follows:

[0040] right and Performing a Fourier transform yields the corresponding two-sided power spectrum:

[0041] ,

[0042] In the formula For the corresponding The two-sided power spectrum, For the corresponding The two-sided power spectrum, The sequence after Fourier transform The number of elements;

[0043] calculate Corresponding one-sided power spectrum and Corresponding one-sided power spectrum :

[0044] ,

[0045] In the formula The number of frequency points is i = 1, 2, 3…floor(N / 2) + 1, where floor represents rounding down. Let i be the i-th Fourier discrete frequency;

[0046] Calculate the one-sided power spectrum The corresponding zero-mean ultrasonic signal power density spectrum and single-sided power spectrum The corresponding zero-mean ultrasonic signal power density spectrum :

[0047] ,

[0048] In the formula For frequency intervals, The signal sampling frequency;

[0049] Zero-mean ultrasonic signal power density spectrum and zero-mean ultrasound signal power density spectrum Perform a normalization transformation to obtain the normalized power density spectrum of the time series before transformation across the entire frequency band. and the normalized power density spectrum of the time series after full-band transformation :

[0050] ,

[0051] Depend on and Obtain the normalized power density spectrum of the time series before transformation corresponding to the peak power density. and the normalized power density spectrum of the transformed time series .

[0052] The beneficial effects of this invention are as follows: This invention monitors the cage rotation speed of aerospace bearings based on the peak power density ratio, which greatly improves the accuracy and robustness of cage rotation speed measurement under high-speed conditions.

[0053] This invention utilizes the characteristic that the peak value of the ultrasonic amplitude power density representing the depth of a typical indentation over which the rolling element passes is most significantly amplified in the upward contraction ultrasonic amplitude spectrum. This overcomes the negative impact of missing rolling element signals on the measurement results and achieves accurate determination of the rolling element passing frequency under high-speed conditions. Compared with time-domain cross-correlation and Fourier transform spectral peak methods, it has higher accuracy and robustness. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of signal monitoring for the ultrasonic monitoring method for the rotational speed of aerospace bearing cages described in this invention;

[0055] Figure 2 This is a schematic diagram of the reflected pulse signal of the bearing when the inner ring rotates at 2000 r / min in a specific embodiment;

[0056] Figure 3 This is a schematic diagram of the reflected pulse signal of the bearing when the inner ring rotates at 10000 r / min in a specific embodiment;

[0057] Figure 4 This is a schematic diagram of the reflected pulse signal of the bearing when the inner ring rotates at 20000 r / min in a specific embodiment;

[0058] Figure 5 yes Figure 2 A magnified view of the time-domain signal of a single reflected pulse.

[0059] Figure 6 This is a schematic diagram of the normalized frequency domain amplitude time series of the ultrasonic reflection signal corresponding to the center frequency of the bearing when the inner ring rotates at 2000 r / min.

[0060] Figure 7 This is a schematic diagram of the normalized frequency domain amplitude time series of the ultrasonic reflection signal corresponding to the center frequency of the bearing when the inner ring rotates at 10000 r / min.

[0061] Figure 8 This is a schematic diagram of the normalized frequency domain amplitude time series of the ultrasonic reflection signal corresponding to the center frequency of the bearing when the inner ring rotates at 20000 r / min.

[0062] Figure 9 This is a schematic diagram of the normalized frequency domain amplitude time series of the ultrasonic reflection signal after the pit shrinkage transformation when the bearing rotates at 2000 r / min on the inner ring.

[0063] Figure 10 This is a schematic diagram of the normalized frequency domain amplitude time series of the ultrasonic reflection signal after the pit shrinkage transformation when the bearing rotates at 10000 r / min on the inner ring.

[0064] Figure 11 This is a schematic diagram of the normalized frequency domain amplitude time series of the ultrasonic reflection signal after the pit shrinkage transformation when the bearing rotates at 20,000 r / min on the inner ring.

[0065] Figure 12 It is the normalized power density spectrum of the ultrasonic reflection signal before the pit shrinkage transformation when the bearing rotates at 2000 r / min in the inner ring.

[0066] Figure 13 It is the normalized power density spectrum of the ultrasonic reflection signal before the pit shrinkage transformation when the bearing rotates at 10000 r / min in the inner ring.

[0067] Figure 14 It is the normalized power density spectrum of the ultrasonic reflection signal before the pit shrinkage transformation when the bearing rotates at 20000 r / min in the inner ring.

[0068] Figure 15 It is the normalized power density spectrum of the ultrasonic reflection signal after the pit shrinkage transformation when the bearing rotates at 2000 r / min in the inner ring.

[0069] Figure 16 It is the normalized power density spectrum of the ultrasonic reflection signal after the pit shrinkage transformation when the bearing rotates at 10000 r / min in the inner ring.

[0070] Figure 17 It is the normalized power density spectrum of the ultrasonic reflection signal after the pit shrinkage transformation when the bearing rotates at 20000 r / min on the inner ring. Detailed Implementation

[0071] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0072] Specific Implementation Method 1: Combination Figure 1 As shown, the present invention provides an ultrasonic monitoring method for the rotational speed of an aerospace bearing cage, comprising:

[0073] An ultrasonic probe was used to acquire the time series of reflected pulse signals as the bearing rolling elements rolled, and Fourier transform was performed to obtain the time series of ultrasonic normalized amplitude at the pulse center frequency.

[0074] The normalized ultrasonic amplitude sequence is processed, and the typical crater signal over which the rolling body rolls is subjected to a contraction transformation to obtain the normalized ultrasonic frequency domain amplitude time sequence after the typical crater contraction transformation at the pulse center frequency.

[0075] The normalized ultrasonic amplitude time series and the normalized ultrasonic frequency domain amplitude time series after typical pit shrinkage transformation are respectively subjected to zero-mean processing to obtain the zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series.

[0076] The zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series are processed to obtain the normalized power density spectrum of the time series before transformation and the normalized power density spectrum of the time series after transformation. The power density peak value and corresponding frequency of the normalized power density spectrum of the time series before transformation are extracted, and the power density peak value of the normalized power density spectrum of the time series after transformation is determined by the frequency.

[0077] Calculate the percentage of the peak power density in the normalized power density spectrum of each time series before transformation, and the percentage of the peak power density in the normalized power density spectrum of each time series after transformation.

[0078] The calculation range of the rolling element passing frequency is determined based on the theoretical value of the rolling element passing frequency. The peak percentages before and after all transformations are filtered to obtain the filtered peak percentages before and after transformations. The peak percentages after transformations are divided by the corresponding points of the filtered peak percentages before transformations, and the frequency corresponding to the maximum value of the peak percentage is taken as the rolling element passing frequency.

[0079] The cage rotational speed is calculated based on the frequency of the rolling elements.

[0080] Furthermore, combined with Figure 1 As shown, using Figure 1 The aerospace bearing cage speed monitoring device shown acquires the time-domain signal of the reflected pulse. The monitoring device includes an ultrasonic probe, an ultrasonic pulse generator and receiver, and a data acquisition, storage, and analysis system. The probe is placed in a water bath directly above the outer ring of the bearing. During testing, the ultrasonic pulse generator and receiver emits excitation pulses at a fixed frequency to the ultrasonic probe. The ultrasonic probe then emits ultrasonic pulses to scan the inner wall of the outer ring of the bearing directly below, generating echo pulses that are received, amplified, and transmitted to the data acquisition, storage, and analysis system. Finally, the signal is analyzed to determine the passing frequency of the rolling elements.

[0081] Time series of reflected pulse signals for:

[0082] ,

[0083] In the formula For time, This is the first reflected pulse signal. This is the second reflected pulse signal. The third reflected pulse signal; 1, 2, 3 represent the order in which the ultrasonic pulse generator and receiver transmit at frequency F. For example, if the transmission frequency F = 50kHz, 2 represents the second pulse received when the interval between the received pulse ∆t = 1 / F and the received pulse 1; Y(t) represents the sequential pulse signal of each single reflected pulse arranged in chronological order.

[0084] Time series of reflected pulse signals Perform a Fourier transform to obtain the amplitude corresponding to the center frequency of each pulse, and arrange them in chronological order to obtain the time series of normalized ultrasonic amplitude at the pulse center frequency. for:

[0085] ,

[0086] In the formula The pulse center frequency, Indicates Fourier transform, Indicates the frequency at the center of the pulse. The value to be taken below, This indicates the amplitude value.

[0087] Because the loss of some rolling element signals at high speeds leads to distortion in cage speed measurement, in order to obtain accurate rolling element passing frequency, a shrinkage transformation is performed on the typical indentations representing the rolling elements in the normalized ultrasonic frequency domain amplitude time history at the above center frequency.

[0088] The normalized ultrasonic frequency domain amplitude time series after typical pit contraction transformation at the pulse center frequency is expressed as follows: :

[0089] ,

[0090] In the formula The threshold for determining whether there are typical dents caused by a rolling element in the normalized amplitude time series of ultrasound. It is an empirical constant that can be determined based on the load conditions of the test bearing and the ultrasonic transmission and reception parameter settings, and its range is generally 0.7~0.85; it represents the normalized ultrasonic frequency domain amplitude time history at the center frequency after the pit shrinkage transformation.

[0091] To facilitate power spectral density analysis, the normalized ultrasonic amplitude time history at the center frequency after the above-mentioned pit shrinkage transformation is subjected to zero-mean processing. The resulting zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series are as follows:

[0092] ,

[0093] In the formula This is a zero-mean normalized ultrasound amplitude time series. This is the transformed zero-mean normalized ultrasonic frequency domain amplitude time series. for The mean, for The mean;

[0094] and The following conditions must be met:

[0095] .

[0096] in This represents the normalized ultrasonic frequency domain amplitude and time history at the center frequency before the pit shrinkage transformation. The mean, This represents the normalized ultrasonic frequency domain amplitude and time history at the center frequency after the pit shrinkage transformation. The mean.

[0097] Subsequently, the normalized power density spectrum of the normalized ultrasonic frequency domain amplitude time history at the zero mean center frequency before and after the above-mentioned pit shrinkage transformation was obtained. Then, the peak value and its corresponding frequency in the normalized power density spectrum of the ultrasonic signal before pit shrinkage were extracted, and the normalized power density peak value of the ultrasonic signal after pit shrinkage transformation was obtained using the frequency.

[0098] The normalized power density spectrum of the time series before transformation is represented as follows: The normalized power density spectrum of the transformed time series is expressed as: Threshold PSD is determined by setting the peak power density. ref ,get:

[0099] ,

[0100] In the formula This is an array of power density peaks in the normalized power density spectrum of the time series before transformation. This is the power density peak array of the normalized power density spectrum of the transformed time series. For frequency The corresponding peak power density of the normalized power density spectrum of the time series before transformation, where k is the number of peaks. for The corresponding power density peak value of the normalized power density spectrum of the transformed time series.

[0101] The method for calculating the percentage of peak power density in the normalized power density spectrum of the time series before transformation and the percentage of peak power density in the normalized power density spectrum of the time series after transformation are as follows:

[0102] ,

[0103] In the formula This is an array of peak percentages before the transformation. This is an array of peak percentages after transformation. The normalized power density peak value of the ultrasonic signal before the pit shrinks; The normalized power density peak value of the ultrasonic signal after the pit shrinks is represented by the peak value.

[0104] Furthermore, in order to accurately obtain the rolling element passing frequency, the calculation range of the rolling element passing frequency is determined by the inner ring speed, which is easy to measure accurately. First, the theoretical value of the rolling element passing frequency is determined based on the measured inner ring speed of the bearing. 10% of the theoretical value is taken as the lower limit of the calculation range of the rolling element passing frequency, and 120% of the theoretical value is taken as the upper limit of the calculation range of the rolling element passing frequency.

[0105] Then, the normalized power density peak ratio array of the ultrasonic signal before and after the pit contraction within the calculation range is obtained, and finally, the frequency corresponding to the peak power density with the largest increase in ratio is obtained:

[0106] The method for calculating the frequency of rolling elements is as follows:

[0107] ,

[0108] In the formula To increase the maximum value of the obtained peak percentage, The rolling element passes through at a frequency that corresponds to the peak power density at which the proportion increases. This is a filtered array of transformed peak percentages. This is a filtered array of peak percentages before transformation. This indicates the dot division operation.

[0109] The frequency obtained above is the rolling element throughput frequency. Based on the relationship between the rolling element throughput frequency and the cage speed, the cage speed is calculated as follows:

[0110] ,

[0111] In the formula To maintain the rack speed, This represents the number of rolling elements.

[0112] Furthermore, the normalized power density spectrum of the time series before transformation and the normalized power density spectrum of the transformed time series The method to obtain it is as follows:

[0113] right and Performing a Fourier transform yields the corresponding two-sided power spectrum:

[0114] ,

[0115] In the formula For the corresponding The two-sided power spectrum, For the corresponding The two-sided power spectrum, The sequence after Fourier transform The number of elements;

[0116] calculate Corresponding one-sided power spectrum and Corresponding one-sided power spectrum :

[0117] ,

[0118] In the formula The number of frequency points is i = 1, 2, 3…floor(N / 2) + 1, where floor represents rounding down. Let i be the i-th Fourier discrete frequency;

[0119] Calculate the one-sided power spectrum The corresponding zero-mean ultrasonic signal power density spectrum and single-sided power spectrum The corresponding zero-mean ultrasonic signal power density spectrum :

[0120] ,

[0121] In the formula For frequency intervals, This is the signal sampling frequency, and its value is consistent with the transmission frequency F.

[0122] Zero-mean ultrasonic signal power density spectrum and zero-mean ultrasound signal power density spectrum Perform a normalization transformation to obtain the normalized power density spectrum of the time series before transformation across the entire frequency band. and the normalized power density spectrum of the time series after full-band transformation ,in Represents frequency Zero-mean ultrasonic signal power density before pit contraction transformation Represents frequency Zero-mean ultrasonic signal power density after indentation shrinkage transformation:

[0123] ,

[0124] Depend on and Obtain the normalized power density spectrum of the time series before transformation corresponding to the peak power density. and the normalized power density spectrum of the transformed time series .

[0125] Example 1:

[0126] Based on the principles of this invention, the following monitoring is performed:

[0127] The following example illustrates the steps of using the method of the present invention: Taking a high-speed roller bearing with an inner ring diameter of 119 mm as an example, the cage speed tests were conducted at inner ring speeds of 2000 r / min (238000 mm·r / min dN value), 10000 r / min (1190000 mm·r / min dN value), and 20000 r / min (2380000 mm·r / min dN value).

[0128] First, the reflected pulse signal of the rolling element of the aerospace bearing passing under the ultrasonic probe at high speed is acquired using signal acquisition equipment. The acquired time-domain signal of a portion of the reflected pulse as the rolling element passes under the ultrasonic probe at high speed is shown below. Figures 2 to 5 As shown, with the increase of rotational speed, the number of pits representing the rolling element passing under the ultrasonic probe increases in the same time period.

[0129] Next, a Fourier transform is performed on the acquired time-domain pulse signal to obtain the frequency-domain amplitude corresponding to the center frequency of the ultrasound probe, resulting in the normalized frequency-domain amplitude time history at the center frequency, as shown below. Figures 6 to 8As shown, similar to the time-domain signal, the number of dents representing the rolling elements passing through increases with increasing rotational speed. At low speeds, the dents are relatively evenly distributed, indicating a constant bearing speed. However, at constant high speeds, the dents, which should be evenly distributed, exhibit significant non-uniformity. This is due to insufficient hardware transmission frequency and high noise, causing the weak reflected signals from unscanned parts of the rolling elements or the scanned areas to be masked by noise.

[0130] Next, the normalized ultrasonic frequency domain amplitude-time history plot at the center frequency obtained above is subjected to a dip-shrink transformation. The transformation rules are as follows: for amplitudes less than 0.8, the value is increased by 0.8, while the others remain unchanged. The transformed normalized ultrasonic frequency domain amplitude-time history plot at the center frequency is shown below. Figures 9 to 11 As shown.

[0131] To facilitate power spectral density analysis, the normalized frequency domain amplitude-time history of the ultrasound signal at the center frequency was subjected to a zero-mean transform. The normalized power density spectrum of the ultrasound signal after the zero-mean transform was then obtained, as shown below. Figures 12 to 14 and Figures 15 to 17 As shown.

[0132] Subsequently, the peak value of the normalized power density and its corresponding frequency were captured, and the proportion of the peak value of the normalized power density was further analyzed. The peak value, proportion and corresponding frequency of the normalized power density at each speed are shown in Table 1 and Table 2.

[0133] Table 1. Peak value, proportion, and frequency of normalized power density of ultrasonic reflection signals before pit shrinkage transformation.

[0134] Table 2. Peak value, proportion, and frequency of normalized power density of ultrasonic reflected signals after pit shrinkage transformation.

[0135]

[0136] To obtain accurate rolling element passing frequencies and determine the calculation range of rolling element frequencies, based on the relationship between the inner ring speed of the roller bearing and the theoretical non-slipping rolling element speed (the number of rolling elements in the selected roller bearing is 28), the calculation range of rolling element frequencies for the corresponding speeds is determined as shown in Table 3:

[0137] Table 3 Frequency Calculation Range at Various Rotational Speeds

[0138]

[0139] The frequency corresponding to the normalized power density peak with the largest change in peak percentage within the frequency calculation range is taken as the actual rolling element passing frequency. Through calculation and comparison, the measured values ​​of the rolling element passing frequency at each speed are determined as shown in Table 4.

[0140] Table 4. Rolling element passing frequency measurements for each rotational speed.

[0141]

[0142] Furthermore, the cage rotation speed was calculated based on the correlation between the rolling element passing frequency and the cage rotation speed. The ultrasonic measurement results under various working conditions and the fiber optic measurement results under conditions with less oil mist are shown in Table 5. The error between the ultrasonic measurement results and the fiber optic measurement results is within 1%, which verifies the accuracy of the method of the present invention.

[0143] Table 5

[0144]

[0145] Example 2:

[0146] Ultrasonic reflected pulses of a certain type of roller bearing (119mm inner diameter, 28 rollers) at an inner ring speed of 3000 r / min were obtained using an aerospace bearing cage speed monitoring device. The results were compared using the method provided in this invention, as well as the cross-correlation method and the maximum Fourier peak value method for ultrasonic cage speed testing. The results are shown in Table 6.

[0147] Table 6

[0148]

[0149] Example 3:

[0150] Ultrasonic reflected pulses of a certain type of roller bearing (119mm inner diameter, 28 rollers) at an inner ring speed of 8000 r / min were obtained using an aerospace bearing cage speed monitoring device. The results were compared using the method provided in this invention, the cross-correlation method for ultrasonic cage speed testing, and the maximum Fourier peak value method. The results are shown in Table 7.

[0151] Table 7

[0152]

[0153] Example 4:

[0154] Ultrasonic reflected pulses of a certain type of roller bearing (119mm inner diameter, 28 rollers) at an inner ring speed of 15000 r / min were obtained using an aerospace bearing cage speed monitoring device. The results were compared using the method provided in this invention, the cross-correlation method for ultrasonic cage speed testing, and the maximum Fourier peak value method. The results are shown in Table 8.

[0155] Table 8

[0156]

[0157] Example 5:

[0158] Ultrasonic reflected pulses of a certain type of ball bearing (100mm inner diameter, 25 balls) at an inner ring speed of 15000 r / min were obtained using an aerospace bearing cage speed monitoring device. The results were compared using the method provided in this invention, the cross-correlation method for ultrasonic cage speed testing, and the maximum Fourier peak value method. The results are shown in Table 9.

[0159] Table 9

[0160]

[0161] Based on the combined results, taking the fiber optic test results under low oil mist conditions as a reference, at low speeds, the method of this invention is basically consistent with the traditional cross-correlation method and the maximum Fourier frequency domain amplitude method. However, at high speeds, the method of this invention overcomes the influence of missing signals of some rolling elements and has greater advantages in terms of test accuracy and robustness.

[0162] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A method for ultrasonic monitoring of the rotational speed of an aerospace bearing cage, characterized in that, include: An ultrasonic probe was used to acquire the time series of reflected pulse signals as the bearing rolling elements rolled, and Fourier transform was performed to obtain the time series of ultrasonic normalized amplitude at the pulse center frequency. The normalized ultrasonic amplitude time series is processed, and the typical crater signal over which the rolling body rolls is subjected to a contraction transformation to obtain the normalized ultrasonic frequency domain amplitude time series after the typical crater contraction transformation at the pulse center frequency. The normalized ultrasonic amplitude time series and the normalized ultrasonic frequency domain amplitude time series after typical pit shrinkage transformation are respectively subjected to zero-mean processing to obtain the zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series. The zero-mean normalized ultrasonic amplitude time series and the transformed zero-mean normalized ultrasonic frequency domain amplitude time series are processed to obtain the normalized power density spectrum of the time series before transformation and the normalized power density spectrum of the time series after transformation. The power density peak value and corresponding frequency of the normalized power density spectrum of the time series before transformation are extracted, and the power density peak value of the normalized power density spectrum of the time series after transformation is determined by the frequency. Calculate the percentage of the peak power density in the normalized power density spectrum of each time series before transformation, and the percentage of the peak power density in the normalized power density spectrum of each time series after transformation. The calculation range of the rolling element passing frequency is determined based on the theoretical value of the rolling element passing frequency. The peak percentages before and after all transformations are filtered to obtain the filtered peak percentages before and after transformations. The peak percentages after transformations are divided by the corresponding points of the filtered peak percentages before transformations, and the frequency corresponding to the maximum value of the peak percentage is taken as the rolling element passing frequency. The cage rotational speed is calculated based on the frequency of the rolling elements. The normalized power density spectrum of the time series before transformation is represented as follows: The normalized power density spectrum of the transformed time series is expressed as: Threshold PSD is determined by setting the peak power density. ref ,get: ， In the formula This is the power density peak array of the normalized power density spectrum of the time series before transformation. This is the power density peak array of the normalized power density spectrum of the transformed time series. For frequency The corresponding peak power density of the normalized power density spectrum of the time series before transformation, where k is the number of peaks. for The peak power density of the corresponding transformed time series normalized power density spectrum; The method for calculating the percentage of peak power density in the normalized power density spectrum of the time series before transformation and the percentage of peak power density in the normalized power density spectrum of the time series after transformation are as follows: ， In the formula This is an array of peak percentages before the transformation. This is an array of peak percentages after transformation.

2. The ultrasonic monitoring method for the rotational speed of aerospace bearing cages according to claim 1, characterized in that, Time series of reflected pulse signals for: ， In the formula For time, This is the first reflected pulse signal. This is the second reflected pulse signal. This is the third reflected pulse signal; Time series of reflected pulse signals Perform a Fourier transform to obtain the normalized amplitude time series of ultrasound at the pulse center frequency. for: ， In the formula The pulse center frequency, Indicates Fourier transform, Indicates the frequency at the center of the pulse. The value below, This indicates the amplitude value.

3. The ultrasonic monitoring method for the rotational speed of aerospace bearing cages according to claim 2, characterized in that, The normalized ultrasonic frequency domain amplitude time series after typical pit contraction transformation at the pulse center frequency is expressed as follows: : ， In the formula The threshold for determining whether there are typical dents caused by a rolling element in the normalized amplitude time series of ultrasound.

4. The ultrasonic monitoring method for the rotational speed of an aerospace bearing cage according to claim 3, characterized in that, The zero-mean normalized ultrasound amplitude time series and the transformed zero-mean normalized ultrasound frequency domain amplitude time series are as follows: ， In the formula This is a zero-mean normalized ultrasound amplitude time series. This is the transformed zero-mean normalized ultrasonic frequency domain amplitude time series. for The mean, for The mean; 。 5. The ultrasonic monitoring method for the rotational speed of an aerospace bearing cage according to claim 4, characterized in that, The theoretical value of the rolling element rolling frequency is determined based on the measured inner ring speed of the bearing. 10% of the theoretical value is taken as the lower limit of the calculation range of the rolling element rolling frequency, and 120% of the theoretical value is taken as the upper limit of the calculation range of the rolling element rolling frequency.

6. The ultrasonic monitoring method for the rotational speed of an aerospace bearing cage according to claim 5, characterized in that, The method for calculating the frequency of rolling elements is as follows: ， In the formula To increase the maximum value of the obtained peak percentage, For the frequency of the rolling element, This is a filtered array of transformed peak percentages. This is a filtered array of peak percentages before transformation.

7. The ultrasonic monitoring method for the rotational speed of an aerospace bearing cage according to claim 6, characterized in that, The method for calculating the cage rotational speed is as follows: ， In the formula To maintain the rack speed, This represents the number of rolling elements.

8. The ultrasonic monitoring method for the rotational speed of an aerospace bearing cage according to claim 7, characterized in that, Normalized power density spectrum of time series before transformation and the normalized power density spectrum of the transformed time series The method to obtain it is as follows: right and Performing a Fourier transform yields the corresponding two-sided power spectrum: ， In the formula For the corresponding The two-sided power spectrum, For the corresponding The two-sided power spectrum, The sequence after Fourier transform The number of elements; calculate Corresponding single-sided power spectrum and Corresponding single-sided power spectrum : ， In the formula Here, i represents the number of frequency points, i = 1, 2, 3… floor(N / 2) + 1, where floor represents rounding down. Let i be the i-th Fourier discrete frequency; Calculate the one-sided power spectrum The corresponding zero-mean ultrasonic signal power density spectrum and single-sided power spectrum The corresponding zero-mean ultrasonic signal power density spectrum : ， In the formula For frequency intervals, The signal sampling frequency; Zero-mean ultrasonic signal power density spectrum and zero-mean ultrasonic signal power density spectrum Perform a normalization transformation to obtain the normalized power density spectrum of the time series before transformation across the entire frequency band. and the normalized power density spectrum of the time series after full-band transformation : ， Depend on and Obtain the normalized power density spectrum of the time series before transformation corresponding to the peak power density. and the normalized power density spectrum of the transformed time series .

Images