A bearing oil film thickness measurement method based on quasi-unipolar pulse ultrasound

Abstract

The application relates to the technical field of bearing oil film thickness measurement, and particularly discloses a bearing oil film thickness measurement method based on quasi-monopole pulse ultrasonic waves, wherein quasi-monopole pulse ultrasonic waves with concentrated time domain energy and extremely wide frequency spectrum coverage are actively synthesized at a sound source end, and then the quasi-monopole pulse ultrasonic waves are shot into a multi-layer medium of bearing steel-oil film-bearing steel; reflected signals are processed, and a full-band reflection coefficient spectrum is extracted; a physical model between a theoretical reflection coefficient spectrum and oil film thickness is established based on a spring model, and a wide-spectrum fitting target function is established; the target function is solved through a nonlinear optimization algorithm to obtain optimal oil film thickness. The application provides an effective method for solving the problems of lack of uniqueness and stability in the bearing oil film thickness measurement inversion process, and simultaneously meets the requirements of high precision, high stability and wide range.

Description

Technical Field

[0001] This invention relates to the field of bearing oil film thickness measurement technology, and in particular to a bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound. Background Technology

[0002] Bearings are core components of critical equipment, and online monitoring of their lubrication status is crucial for ensuring the reliability, lifespan, and safety of equipment operation. Lubricating oil film thickness, as the most fundamental physical indicator reflecting the evolution of lubrication status from boundary lubrication and mixed lubrication to full-film hydrodynamic lubrication, is of great significance for achieving bearing health management. This is especially true for rolling bearings, where the oil film thickness is typically only 0.1-5 μm, placing extremely high demands on the resolution and accuracy of measurement techniques.

[0003] Currently, the main methods for measuring bearing oil film thickness include optical, electrical, and ultrasonic methods. Optical methods (such as interferometry and fluorescence methods) offer high accuracy, but often require at least one friction pair to be made of a transparent material, making in-situ online monitoring difficult on all-metal bearings. Electrical methods (such as capacitance and resistance methods) are easy to integrate, but they reflect the average electrical parameters across the entire bearing pair, resulting in limited accuracy, and are prone to electrical breakdown under thin oil films. Both of these methods face significant challenges in terms of quantitative accuracy, spatial resolution, and reliability under complex operating conditions.

[0004] Ultrasonic methods, with their advantages of being non-invasive, having strong penetrating power, wide applicability, and easy online integration, have shown promising prospects for engineering applications. Existing methods for measuring bearing oil film thickness based on ultrasonic principles mainly include the phase method and the spring model method. The phase method utilizes the functional relationship between the phase lag caused by ultrasonic waves reflecting off the upper and lower interfaces of the oil film and the film thickness. It is sensitive to changes in micron-level thin oil films, but its accuracy is easily affected by material sound velocity, temperature fluctuations, and system noise, resulting in insufficient stability. The spring model method treats the viscous oil film as a massless "spring," and by constructing a physical model with parameters such as the reflection coefficient spectrum and film thickness, it can achieve sub-micron-level high-precision measurements. However, the accuracy, uniqueness, and effective measurement range of this method are highly dependent on the effective bandwidth of the incident ultrasonic signal. Traditional single-frequency or narrow-band ultrasonic excitation sources generate limited acoustic wave information, and the spectral characteristics of reflections at the oil film interface are singular, lacking sufficient constraints. This makes it difficult to meet the high-precision measurement requirements for a wide range (0.1-5μm) of thin oil films in terms of inversion accuracy and stability.

[0005] To address the aforementioned challenges, there is an urgent need for a high-precision bearing oil film thickness measurement method that can actively synthesize a broadband ultrasonic source and combine it with broadband spectrum fitting and inversion, in order to overcome the contradiction between measurement accuracy, stability, and range in existing technologies. Summary of the Invention

[0006] The purpose of this invention is to provide a high-precision measurement method for bearing oil film thickness based on quasi-monopolar pulse ultrasound. By actively synthesizing quasi-monopolar pulse ultrasound with concentrated time-domain energy and extremely wide spectral coverage at the sound source end, and constructing an inversion method that integrates a spring model and wide-spectrum fitting, a large number of spectral data points are used to form strong constraints, thereby solving the problems of insufficient uniqueness and stability in the inversion process, and achieving high-precision measurement of bearing oil film thickness in a wide range of 0.1-5μm.

[0007] To achieve the above objectives, this invention provides a method for measuring bearing oil film thickness based on quasi-monopolar pulse ultrasound, comprising the following steps: S1. The dual-band transducer emits quasi-monopolar pulse ultrasound, which is perpendicularly incident on the bearing steel-oil film-bearing steel multilayer medium. The dual-band transducer receives the reflected echo signal of the quasi-monopolar pulse ultrasound reflected in the bearing steel-oil film-bearing steel multilayer medium. S2. Perform preprocessing on the reflected echo signal, including denoising, filtering, and normalization, and extract the full-band reflection coefficient spectrum; S3. Based on the spring model, establish a physical model between the theoretical reflection coefficient spectrum and the oil film thickness, and establish a wide-spectrum fitting objective function; S4. Solve the objective function iteratively using a nonlinear optimization algorithm to obtain the optimal oil film thickness.

[0008] Furthermore, in step S1, the structure of the dual-band transducer, from bottom to top, consists of a backing layer, a low-frequency piezoelectric layer, an acoustic filter layer, a high-frequency piezoelectric layer, a first matching layer, and a second matching layer. The method for synthesizing quasi-monopolar pulse ultrasound is as follows: the low-frequency piezoelectric layer and the high-frequency piezoelectric layer in the dual-band transducer receive synchronous electrical excitation signals respectively, and the amplitude and phase relationship of the two excitation signals are adjusted so that the low-frequency ultrasonic waves generated by the low-frequency piezoelectric layer and the high-frequency ultrasonic waves generated by the high-frequency piezoelectric layer undergo time-domain interference superposition in the propagation path, thereby synthesizing quasi-monopolar pulse ultrasound with a prominent main lobe and suppressed side lobes.

[0009] Furthermore, in step S1, in the dual-band transducer, the center frequency of the low-frequency piezoelectric layer is 1MHz-10MHz, the center frequency of the high-frequency piezoelectric layer is 3MHz-50MHz, and the center frequency of the high-frequency piezoelectric layer is 3-5 times that of the center frequency of the low-frequency piezoelectric layer.

[0010] Furthermore, the ratio of the acoustic impedance of the backing layer to the acoustic impedance of the low-frequency piezoelectric layer of the dual-band transducer ranges from 0.4:1 to 0.8:1. The thickness of the first and second matching layers of the dual-band transducer is 0.1-0.3 times the wavelength of the high-frequency ultrasonic wave, the thickness of the acoustic filter layer is 0.1-0.5 times the wavelength of the high-frequency ultrasonic wave, the thickness of the low-frequency matching layer is 0.1-1 times the wavelength of the low-frequency ultrasonic wave, and the thickness of the high-frequency matching layer is 0.1-1 times the wavelength of the high-frequency ultrasonic wave.

[0011] Furthermore, the quasi-monopolar pulse ultrasound has a -6dB relative bandwidth >140% and a main lobe to side lobe amplitude ratio greater than 3.5:1.

[0012] Furthermore, in step S2, the preprocessing specifically includes: The reflected echo signal is denoised using a digital bandpass filter, the passband frequency range of which covers the spectral width of the quasi-monopolar pulse. A wavelet threshold denoising method is used to perform secondary denoising on the filtered signal to eliminate environmental noise and system electrical noise. Finally, a maximum amplitude normalization method is used to normalize the amplitude of the denoised signal to the [-1, 1] interval, eliminating the influence of the sensitivity difference between the two-band transducers on the absolute value of the reflection coefficient.

[0013] Furthermore, in step S2, the extraction of the full-band reflection coefficient spectrum specifically involves: extracting a high-fidelity reflection coefficient spectrum from the reflected echo signal using a deconvolution method.

[0014] Furthermore, step S3 specifically includes: Based on the spring model, the oil film is equivalent to an elastic interface, and the correspondence between reflectivity and the mechanical property parameters of the oil film is established. Using oil film thickness as the parameter to be determined, a wide-spectrum fitting objective function is constructed based on the difference between the full-band reflection coefficient spectrum and the theoretical reflection coefficient spectrum.

[0015] Furthermore, the expression for the correspondence between the reflectivity and the oil film mechanical property parameters is as follows: ; in, Reflectivity; The acoustic impedance of the first layer of medium; The acoustic impedance of the third layer of medium; ω is the angular frequency of the ultrasound. This is the stiffness coefficient of the oil film layer; The stiffness coefficient of the oil film layer is: ; in, The density of the oil film medium; The velocity of sound in the oil film medium; Oil film thickness; When the oil film thickness has a linear relationship with the reflection coefficient, and the first and third layers of media are identical and have the same acoustic impedance. When the oil film thickness is calculated, the following can be determined: ; in, is the incident frequency.

[0016] Furthermore, the expression for the wide-spectrum fitting objective function is: ; in, Let be the objective function value, and let be the optimization objective. Take the minimum value. The parameters to be inverted; Number of frequency points; Frequency point number; As a weighting factor; The full-band reflection coefficient spectrum; This is the theoretical reflection coefficient spectrum; For Tikhonov regularization terms, For prior estimates of parameters including thickness and viscosity, The regularization coefficient is . It is the square of the L2 norm.

[0017] The present invention employs the above-mentioned method for measuring bearing oil film thickness based on quasi-monopolar pulse ultrasound, and its beneficial effects are as follows: (1) This invention actively synthesizes quasi-monopolar pulse ultrasound with concentrated time-domain energy and extremely wide spectral coverage (-6 dB relative bandwidth > 140%) at the sound source end, providing rich spectral information and strong constraints for inversion, and providing an effective method to solve the uniqueness and stability problems caused by insufficient information in traditional narrowband ultrasound inversion; (2) This invention provides a nonlinear optimization inversion method that integrates spring model and wide spectrum fitting, and introduces deconvolution and Tikhonov regularization strategy, which effectively improves the robustness and measurement accuracy of the inversion algorithm and achieves high-precision measurement in a wide thickness range of 0.1-5μm; (3) The relative error of the inversion in the core measurement range of 0.25-2μm is generally less than 1.5% (e.g., the error is only 0.03% at a thickness of 1μm), and the goodness of fit R 2 With a value greater than 0.95, even under extreme thickness conditions of 0.1μm and 5μm, the relative errors are only 3.9% and 6.35% respectively, breaking through the contradiction between accuracy and range in traditional ultrasonic thickness measurement methods; (4) This invention is applicable to all-metal bearing structures, without changing the original bearing design, and can realize in-situ, online, non-invasive high-precision monitoring. Attached Figure Description

[0018] Figure 1 This is a flowchart of a bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the dual-band transducer structure and transmission line equivalent model of a bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to an embodiment of the present invention; wherein, (a) is the dual-band transducer structure, and (b) is the dual-band transducer transmission line model. Figure 3 This is a quasi-unipolar pulse ultrasound synthesized waveform and spectrum diagram of a bearing oil film thickness measurement method based on quasi-unipolar pulse ultrasound according to an embodiment of the present invention; wherein, (a) is a schematic diagram of quasi-unipolar pulse ultrasound synthesized by a dual-band transducer structure, and (b) is a spectrum diagram of quasi-unipolar pulse ultrasound signal. Figure 4 This is a figure showing the fitting and inversion results of the reflection coefficient spectrum of a bearing oil film thickness based on quasi-monopolar pulse ultrasound according to an embodiment of the present invention; where (a) is the objective function as a function of thickness, (b) is the amplitude spectrum fitting result, and (c) is the residual distribution. Figure 5 This is a graph showing the variation of inversion parameters for 0.1-5μm oil film thickness in an embodiment of the present invention, based on a quasi-monopolar pulse ultrasound method for measuring bearing oil film thickness; where (a) represents the thickness inversion result and accuracy, (b) represents the model fitting quality, and (c) represents the relative error analysis. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0020] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for measuring bearing oil film thickness based on quasi-monopolar pulse ultrasound, comprising the following steps: S1. The dual-band transducer emits quasi-monopolar pulse ultrasound, which is perpendicularly incident on the bearing steel-oil film-bearing steel multilayer medium. The dual-band transducer receives the reflected echo signal of the quasi-monopolar pulse ultrasound reflected in the bearing steel-oil film-bearing steel multilayer medium. like Figure 2 As shown, Figure 2(a) shows the structure of a dual-band transducer, consisting of a backing layer 1, a low-frequency piezoelectric layer 2, an acoustic filter layer 3, a high-frequency piezoelectric layer 4, a first matching layer 5, and a second matching layer 6, from bottom to top. The backing layer 1 is used to absorb back-radiated ultrasonic waves and suppress pulse tailing. The low-frequency piezoelectric layer 2 (center frequency of 1MHz-10MHz, 3MHz in this embodiment) generates low-frequency ultrasonic waves. The acoustic filter layer 3 is used to isolate acoustic crosstalk between the low-frequency and high-frequency channels. The high-frequency piezoelectric layer 4 (center frequency of 3MHz-50MHz, 10MHz in this embodiment) generates high-frequency ultrasonic waves, and the center frequency of the high-frequency piezoelectric layer is 3-5 times that of the center frequency of the low-frequency piezoelectric layer. The first matching layer 5 and the second matching layer 6 are used to achieve acoustic impedance matching between the piezoelectric layer and the load medium, thereby improving energy transmission efficiency. The thickness of the first and second matching layers of the dual-band transducer is 0.1-0.3 times the wavelength of the high-frequency ultrasonic wave, the thickness of the acoustic filter layer is 0.1-0.5 times the wavelength of the high-frequency ultrasonic wave, the thickness of the low-frequency matching layer is 0.1-1 times the wavelength of the low-frequency ultrasonic wave, and the thickness of the high-frequency matching layer is 0.1-1 times the wavelength of the high-frequency ultrasonic wave.

[0021] Figure 2 (b) is the equivalent transmission line model corresponding to the dual-band transducer, where each layer is considered as a transmission line segment with characteristic impedance and propagation time, facilitating the analysis of the propagation, reflection, and transmission characteristics of ultrasound. This transducer synthesizes quasi-monopolar pulse ultrasound in the propagation path through dual-band synchronous excitation.

[0022] The synthesis method of quasi-monopolar pulse ultrasound is as follows: In a dual-band transducer, the low-frequency piezoelectric layer and the high-frequency piezoelectric layer respectively receive synchronous electrical excitation signals. The amplitude and phase relationship of the two excitation signals are adjusted so that the low-frequency ultrasonic waves generated by the low-frequency piezoelectric layer and the high-frequency ultrasonic waves generated by the high-frequency piezoelectric layer undergo temporal interference superposition in the propagation path, thereby synthesizing a quasi-monopolar pulse ultrasound with a prominent main lobe and suppressed side lobes. The ratio of the excitation voltage amplitudes of the low-frequency piezoelectric layer to the high-frequency piezoelectric layer is 1-6.

[0023] In this embodiment, the backing layer 1 has an acoustic impedance of 10.5 MRayl and a thickness of 1.085 mm; the low-frequency piezoelectric layer 2 uses PZT-5H piezoelectric ceramic material with a center frequency of 3 MHz and a thickness of 0.53 mm; the acoustic filter layer 3 uses polymer material with an acoustic impedance of 2.4 MRayl and a thickness of 0.053 mm; the high-frequency piezoelectric layer 4 uses PZT-5H piezoelectric ceramic material with a center frequency of 10 MHz and a thickness of 0.16 mm; the first matching layer 5 has an acoustic impedance of 5.4 MRayl and a thickness of 0.075 mm; and the second matching layer 6 has an acoustic impedance of 2.4 MRayl and a thickness of 0.060 mm. The ratio of the acoustic impedance of the backing layer 1 to that of the low-frequency piezoelectric layer 2 (14.4 MRayl) is 0.73:1, which meets the range of 0.4:1-0.8:1.

[0024] During operation, signal generator channel 1 outputs a low-frequency electrical excitation signal with an amplitude of 70V and a center frequency of 3MHz to low-frequency piezoelectric layer 2, and channel 2 outputs a high-frequency electrical excitation signal with an amplitude of 13V and a center frequency of 10MHz to high-frequency piezoelectric layer 4. The two signals are triggered synchronously and are phase-aligned. Low-frequency piezoelectric layer 2 generates low-frequency ultrasonic waves, and high-frequency piezoelectric layer 4 generates high-frequency ultrasonic waves. The forward waves of the two superimpose in the propagation path to synthesize quasi-monopolar pulse ultrasonic waves.

[0025] like Figure 3 As shown, Figure 3 (a) shows the time-domain waveforms of low-frequency ultrasound, high-frequency ultrasound, and the superposition of the two quasi-monopolar pulse ultrasound. As can be seen from the figure, the low-frequency ultrasound mainly provides the pulse body energy, while the high-frequency ultrasound is used to correct the pulse waveform details. When the two signals are superimposed according to a predetermined amplitude ratio and phase relationship, a quasi-monopolar pulse ultrasound with significantly enhanced main lobe and significantly reduced negative tail can be formed. Figure 3 (b) shows the corresponding spectral distribution; it can be seen that the synthesized quasi-monopolar pulse ultrasound simultaneously covers both low and high frequency bands, and its spectral width is significantly greater than that of a single-frequency ultrasound. Since the wide bandwidth contains a large number of frequency sampling points, it can provide richer constraint information for subsequent reflection coefficient inversion. In this embodiment, the -6dB relative bandwidth of the quasi-monopolar pulse ultrasound exceeds 140%, and the amplitude ratio of the main lobe to the side lobe is greater than 3.5:1. Compared with traditional narrowband ultrasound signals, its spectral coverage is significantly expanded, which can effectively improve the parameter identifiability and solution stability in the oil film thickness inversion process.

[0026] S2. Perform preprocessing on the received reflected echo signal, including denoising, filtering, and normalization, and extract the full-band reflection coefficient spectrum using deconvolution and Tikhonov regularization methods. The preprocessing specifically involves: using a digital bandpass filter to denoise the reflected echo signal, wherein the passband frequency range of the digital bandpass filter covers the spectral width of the quasi-monopolar pulse; using a wavelet threshold denoising method to perform secondary denoising on the filtered signal to eliminate environmental noise and system electrical noise; and using a maximum amplitude normalization method to normalize the amplitude of the denoised signal to the [-1,1] interval to eliminate the influence of the sensitivity difference of the dual-band transducers on the absolute value of the reflection coefficient.

[0027] The extraction of the full-band reflection coefficient spectrum specifically involves using the deconvolution method to extract a high-fidelity reflection coefficient spectrum from the reflected echo signal.

[0028] S3. Establish a physical model based on the spring model to correlate the theoretical reflection coefficient spectrum with the oil film thickness, and establish a wide-spectrum fitting objective function; specifically including: S31. Based on the spring model, the oil film is equivalent to a massless spring (elastic interface), and the correspondence between reflectivity and the mechanical property parameters of the oil film is established. The expression for the correspondence between reflectivity and oil film mechanical properties is as follows: ; in, Reflectivity; The acoustic impedance of the first layer of medium; The acoustic impedance of the third layer of medium; ω is the angular frequency of the ultrasound. This is the stiffness coefficient of the oil film layer; The stiffness coefficient of the oil film layer is: ; in, The density of the oil film medium; The velocity of sound in the oil film medium; Oil film thickness; When the oil film thickness has a linear relationship with the reflection coefficient, and the first and third layers of media are identical and have the same acoustic impedance. When the oil film thickness is calculated, the following can be determined: ; in, is the incident frequency.

[0029] S32. Using the oil film thickness as the parameter to be determined, construct a wide-spectrum fitting objective function based on the difference between the full-band reflection coefficient spectrum and the theoretical reflection coefficient spectrum.

[0030] The expression for the wide-spectrum fitting objective function is: ; in, Let be the objective function value, and let be the optimization objective. Take the minimum value. The parameters to be inverted; Number of frequency points; Frequency point number; As a weighting factor; The full-band reflection coefficient spectrum; This is the theoretical reflection coefficient spectrum; For Tikhonov regularization terms, For prior estimates of parameters including thickness and viscosity, The regularization coefficient is . It is the square of the L2 norm.

[0031] S4. The objective function is solved iteratively using a nonlinear optimization algorithm to achieve a globally optimal fit between the theoretical reflection coefficient spectrum and the measured full-band reflection coefficient spectrum across the entire frequency band. Through iteration... Minimizing this will ultimately output the optimal thickness solution.

[0032] like Figure 4 As shown, Figure 4 (b) contains two reflection coefficient spectrum curves: one is the "measured full-band reflection coefficient spectrum" (scatter plot) extracted from the measured echo using deconvolution; the other is the "optimal theoretical reflection coefficient spectrum" (fitting line) obtained by iteratively optimizing the physical model. The two curves show good consistency within the 2MHz-18MHz frequency range. During the inversion process, oil film thickness was used as the parameter to be optimized. A nonlinear optimization algorithm was used to continuously adjust the theoretical model parameters to minimize the error between the theoretical and measured spectra. After iterative solving, the optimal thickness value obtained was 1.0003μm, highly consistent with the set value of 1μm, with a relative error as low as 0.03%, and a goodness of fit R0. 2 The accuracy reached 0.9867, and the standard deviation of the thickness estimation was only 7.9 nm. This result proves that the present invention utilizes a large number of data points in an ultrawide spectrum provided by quasi-monopolar pulse ultrasound to form a strong constraint, enabling extremely high-precision measurements at the sub-micron or even nanometer level.

[0033] Example 2: To verify the measurement accuracy and stability of the present invention over a wide thickness range, this example follows the same method and steps as Example 1, and performs inversion simulations for 10 different thickness conditions with oil film thickness settings of 0.1μm, 0.25μm, 0.5μm, 0.75μm, 1μm, 1.5μm, 2μm, 3μm, 4μm, and 5μm.

[0034] like Figure 5 As shown, the combined inversion thickness and goodness of fit R 2 Analysis was conducted on parameters such as relative error: Within the core measurement range of 0.25-2μm, the inversion relative error was generally below 1.5%, and the goodness of fit R0 was [value missing]. 2 All values ​​are greater than 0.95, demonstrating excellent measurement accuracy and reliability. The relative error is minimal at a thickness of 1 μm, at only 0.03%. Even under extreme thickness conditions (0.1 μm and 5 μm), this method maintains stable inversion capability, with relative errors of 3.9% and 6.35%, respectively. These results significantly overcome the inherent contradiction between accuracy and measurement range in traditional ultrasonic thickness measurement methods.

[0035] Therefore, this invention employs the aforementioned method for measuring bearing oil film thickness based on quasi-monopolar pulse ultrasound. It actively synthesizes quasi-monopolar pulse ultrasound with concentrated time-domain energy and extremely wide spectral coverage (-6 dB relative bandwidth > 140%) at the sound source end, providing rich spectral information and strong constraints for inversion. This provides an effective method to solve the uniqueness and stability problems caused by insufficient information in traditional narrowband ultrasound inversion. A nonlinear optimization inversion method integrating a spring model and wide-spectrum fitting is provided, and deconvolution and Tikhonov regularization strategies are introduced to effectively improve the robustness and measurement accuracy of the inversion algorithm, achieving high-precision measurement within a wide thickness range of 0.1-5 μm. Within the core measurement range of 0.25-2 μm, the relative error of the inversion is generally less than 1.5% (e.g., only 0.03% error at a thickness of 1 μm), and the goodness of fit R0 is high. 2 With a value greater than 0.95, even under extreme thickness conditions of 0.1μm and 5μm, the relative errors are only 3.9% and 6.35% respectively, breaking through the contradiction between accuracy and range in traditional ultrasonic thickness measurement methods. This invention is applicable to all-metal bearing structures, without changing the original bearing design, and can realize in-situ, online, non-invasive high-precision monitoring.

[0036] It is worth noting that all contents not described in detail in this invention are existing technologies and are well known to those skilled in the art.

[0037] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for measuring bearing oil film thickness based on quasi-monopolar pulse ultrasound, characterized in that, Includes the following steps: S1. The dual-band transducer emits quasi-monopolar pulse ultrasound, which is perpendicularly incident on the bearing steel-oil film-bearing steel multilayer medium. The dual-band transducer receives the reflected echo signal of the quasi-monopolar pulse ultrasound reflected in the bearing steel-oil film-bearing steel multilayer medium. S2. Perform preprocessing on the reflected echo signal, including denoising, filtering, and normalization, and extract the full-band reflection coefficient spectrum; S3. Based on the spring model, establish a physical model between the theoretical reflection coefficient spectrum and the oil film thickness, and establish a wide-spectrum fitting objective function; S4. Solve the objective function iteratively using a nonlinear optimization algorithm to obtain the optimal oil film thickness.

2. The bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to claim 1, characterized in that, In step S1, the structure of the dual-band transducer from bottom to top consists of a backing layer, a low-frequency piezoelectric layer, an acoustic filter layer, a high-frequency piezoelectric layer, a first matching layer, and a second matching layer. The method for synthesizing quasi-monopolar pulse ultrasound is as follows: the low-frequency piezoelectric layer and the high-frequency piezoelectric layer in the dual-band transducer receive synchronous electrical excitation signals respectively, and the amplitude and phase relationship of the two excitation signals are adjusted so that the low-frequency ultrasonic waves generated by the low-frequency piezoelectric layer and the high-frequency ultrasonic waves generated by the high-frequency piezoelectric layer undergo time-domain interference superposition in the propagation path, thereby synthesizing quasi-monopolar pulse ultrasound with a prominent main lobe and suppressed side lobes.

3. The bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to claim 2, characterized in that, In step S1, in the dual-band transducer, the center frequency of the low-frequency piezoelectric layer is 1MHz-10MHz, the center frequency of the high-frequency piezoelectric layer is 3MHz-50MHz, and the center frequency of the high-frequency piezoelectric layer is 3-5 times that of the center frequency of the low-frequency piezoelectric layer.

4. The bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to claim 2, characterized in that, The ratio of the acoustic impedance of the backing layer to the acoustic impedance of the low-frequency piezoelectric layer of the dual-band transducer ranges from 0.4:1 to 0.8:

1. The thickness of the first and second matching layers of the dual-band transducer is 0.1-0.3 times the wavelength of the high-frequency ultrasonic wave, the thickness of the acoustic filter layer is 0.1-0.5 times the wavelength of the high-frequency ultrasonic wave, the thickness of the low-frequency matching layer is 0.1-1 times the wavelength of the low-frequency ultrasonic wave, and the thickness of the high-frequency matching layer is 0.1-1 times the wavelength of the high-frequency ultrasonic wave.

5. The bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to claim 2, characterized in that, The quasi-monopolar pulse ultrasound has a -6dB relative bandwidth >140% and a main lobe to side lobe amplitude ratio greater than 3.5:

1.

6. The bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to claim 1, characterized in that, In step S2, the preprocessing specifically includes: The reflected echo signal is denoised using a digital bandpass filter, the passband frequency range of which covers the spectral width of the quasi-monopolar pulse. A wavelet threshold denoising method is used to perform secondary denoising on the filtered signal to eliminate environmental noise and system electrical noise. Finally, a maximum amplitude normalization method is used to normalize the amplitude of the denoised signal to the [-1, 1] interval, eliminating the influence of the sensitivity difference between the two-band transducers on the absolute value of the reflection coefficient.

7. The bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to claim 1, characterized in that, In step S2, the extraction of the full-band reflection coefficient spectrum specifically involves: extracting a high-fidelity reflection coefficient spectrum from the reflected echo signal using a deconvolution method.

8. The bearing oil film thickness measurement method based on quasi-monopolar pulse ultrasound according to claim 1, characterized in that, Step S3 specifically includes: Based on the spring model, the oil film is equivalent to an elastic interface, and the correspondence between reflectivity and the mechanical property parameters of the oil film is established. Using oil film thickness as the parameter to be determined, a wide-spectrum fitting objective function is constructed based on the difference between the full-band reflection coefficient spectrum and the theoretical reflection coefficient spectrum.

9. A method for measuring bearing oil film thickness based on quasi-monopolar pulse ultrasound according to claim 8, characterized in that, The expression for the correspondence between reflectivity and oil film mechanical property parameters is as follows: ； in, Reflectivity; The acoustic impedance of the first layer of medium; The acoustic impedance of the third layer of medium; ω is the angular frequency of the ultrasound. This is the stiffness coefficient of the oil film layer; The stiffness coefficient of the oil film layer is: ； in, The density of the oil film medium; The velocity of sound in the oil film medium; Oil film thickness; When the oil film thickness has a linear relationship with the reflection coefficient, and the first and third layers of media are identical and have the same acoustic impedance. When the oil film thickness is calculated, the following can be determined: ； in, is the incident frequency.

10. A method for measuring bearing oil film thickness based on quasi-monopolar pulse ultrasound according to claim 8, characterized in that, The expression for the wide-spectrum fitting objective function is: ； in, Let be the objective function value, and let be the optimization objective. Take the minimum value. The parameters to be inverted; Number of frequency points; Frequency point number; As a weighting factor; The full-band reflection coefficient spectrum; This is the theoretical reflection coefficient spectrum; For Tikhonov regularization terms, For prior estimates of parameters including thickness and viscosity, The regularization coefficient is . It is the square of the L2 norm.

Images