A method and system for ultrasonic decoupling monitoring of composite loads on intelligent bolts

Abstract

This invention discloses an intelligent bolt composite load ultrasonic decoupling monitoring method and system, belonging to the field of intelligent fastener composite load technology. The invention uses four fan-shaped thin-film piezoelectric ultrasonic transducer regions arranged circumferentially along the bolt head end face to collect ultrasonic echo signals from different directions. It extracts features such as ultrasonic time-of-flight variation, echo amplitude, echo energy, phase, or spectrum for each channel, and constructs axial force-sensitive mean features, bending moment-sensitive opposing differential features, shear force-sensitive neighborhood asymmetry features, and amplitude / energy perturbation features. Finally, it outputs the bolt axial force, shear force in two orthogonal directions, and bending moment in two orthogonal directions through a composite load decoupling model. This invention obtains multi-channel ultrasonic responses in different circumferential directions through a four-sector thin-film piezoelectric ultrasonic transducer array, overcoming the limitation of a single ultrasonic channel only reflecting local or axial stress states, and improving the stability of axial force monitoring.

Description

Technical Field

[0001] This invention belongs to the field of intelligent fastener composite load technology, and specifically relates to an ultrasonic decoupling monitoring method and system for intelligent bolt composite load. Background Technology

[0002] With the development of smart bolts and structural health monitoring technologies, existing solutions have attempted to integrate piezoelectric ultrasonic transducers with bolts. For example, piezoelectric ceramic sheets can be attached to the bolt end face, or thin-film piezoelectric ultrasonic transducers can be fabricated at the bolt head or tail end face. Thin-film piezoelectric ultrasonic transducers offer advantages such as small size, high integration, no need for coupling agents, minimal disturbance to existing connection structures, and suitability for long-term online monitoring, providing a new technological path for multi-state monitoring of smart bolts.

[0003] However, most existing ultrasonic smart bolt technologies still primarily focus on measuring axial preload or axial load, i.e., inverting the axial load through a single ultrasonic propagation path or a single change in time-of-flight. For most ideal assembly conditions, this method can reflect the axial stress state of the bolt relatively well; however, under real engineering service conditions, bolts often do not only bear a single axial load. When the connection interface experiences micro-slippage, lateral impact, or eccentric loading, the bolt shank will undergo lateral shear deformation and bending deformation, and the stress distribution across the bolt cross-section will no longer be uniform. If a single change in axial time-of-flight or a single change in flight time is still used to determine the bolt condition, it may be unable to distinguish the changes in ultrasonic response caused by axial preload attenuation, lateral shearing, eccentric bending, and multiple coupled loads.

[0004] Some existing technical solutions attempt to measure bolt shear force, lateral load magnitude and direction, or synchronous changes in axial force and shear force using multi-transducer, multi-point, or multi-path ultrasonic methods. These solutions have expanded the force dimensions of ultrasonic bolt monitoring to some extent. However, from the perspective of composite load identification, existing solutions still have the following shortcomings: First, some solutions mainly target axial load and shear load, lacking synchronous identification of bending moment components; second, some solutions mainly focus on the magnitude and direction of lateral load, failing to form a multi-component decoupled output oriented towards axial force, bidirectional shear force, and bidirectional bending moment; third, some solutions still use a single acoustic time difference or a single calibration coefficient for load conversion, making it difficult to handle the cross-sensitivity and coupled response between axial force, shear force, and bending moment; fourth, although some solutions use array-type or multi-region transducer structures, they do not fully utilize the characteristics such as mean, counter-differentiation, neighborhood asymmetry, amplitude difference, and energy difference between different sector responses to construct a load-sensitive feature system.

[0005] From a mechanical perspective, axial load, shear force, and bending moment affect the ultrasonic response of bolts in different ways. Axial load typically causes the stress on the bolt cross-section to change approximately uniformly, resulting in consistent changes in ultrasonic time-of-flight across different circumferential orientations. Bending moment causes tension on one side of the bolt cross-section and compression on the other, leading to differences in ultrasonic responses in opposing orientations. Shear force can cause lateral deformation, changes in local contact state, micro-slip at the interface, attenuation of echo amplitude, changes in energy distribution, or enhanced mode conversion, making its response more complex than that of axial load and bending moment. Therefore, it is difficult to simultaneously decouple axial force, bidirectional shear force, and bidirectional bending moment using only a single channel transit time (ToF), a single opposing differential, or a single shear force calibration coefficient. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent ultrasonic decoupling monitoring method and system for composite loads on bolts, addressing the aforementioned problems. This aims to improve the situation where a single ultrasonic channel or a single time-difference characteristic is insufficient to distinguish different load components when critical bolts are simultaneously subjected to axial preload, transverse shear force, eccentric load, and bending moment coupling during actual service.

[0007] The technical solution adopted in this invention is as follows: an intelligent bolt composite load ultrasonic decoupling monitoring system, the system comprising the following components: Smart bolts are used as monitored connection fasteners. A four-sector thin-film piezoelectric ultrasonic transducer array is disposed on the end face of the head of a smart bolt, including a first sector, a second sector, a third sector, and a fourth sector arranged circumferentially, for obtaining ultrasonic response signals in different circumferential orientations respectively. The multi-channel ultrasonic excitation and reception module is used to independently excite, independently receive, or time-division multiplex the four-sector thin-film piezoelectric ultrasonic transducer region and obtain ultrasonic echo signals from four channels. The multi-channel ultrasound feature extraction module is used to extract ultrasound time-of-flight variation, echo amplitude, echo energy, phase, spectral features, or mode conversion features from ultrasound echo signals from each channel. The composite feature construction module is used to construct axial force-sensitive mean features, bending moment-sensitive opposing difference features, shear force-sensitive neighborhood asymmetric features, and amplitude / energy perturbation features based on the multi-channel ultrasonic features of four sectors, and form a composite ultrasonic feature vector. The composite load decoupling module is used to input the composite ultrasonic feature vector into the composite load decoupling model and output the axial force, bidirectional shear force, and bidirectional bending moment of the smart bolt. The bidirectional shear force includes shear force in two orthogonal directions, and the bidirectional bending moment includes bending moment in two orthogonal directions. The bidirectional shear force includes shear force in the first direction and shear force in the second direction; the bidirectional bending moment includes bending moment in the first direction and bending moment in the second direction. The calibration parameter storage module is used to store composite load decoupling matrices, bias vectors, cross sensitivity coefficients, or model parameters. The load status output module is used to output the composite load status of the smart bolt, including at least one of axial force, shear force in the first direction, shear force in the second direction, bending moment in the first direction, and bending moment in the second direction.

[0008] Furthermore, the four-sector thin-film piezoelectric ultrasonic transducer array is set in the central region of the hexagonal head end face of the smart bolt or in the annular region surrounding the central region of the end face; the four sector-shaped thin-film piezoelectric ultrasonic transducer regions are evenly arranged along the circumference and are denoted as A, B, C, and D, corresponding to 0°, 90°, 180°, and 270° azimuths, respectively; the first sector A and the third sector C form a pair of opposing sectors, and the second sector B and the fourth sector D form another pair of opposing sectors.

[0009] It should be noted that each sector is independently excited and received, and time-division multiplexing is used to achieve integrated transmission and reception within the same sector.

[0010] Furthermore, let the changes in ultrasonic time-of-flight for the four sectors be as follows:

[0011] The corresponding echo amplitudes are as follows:

[0012] The corresponding echo energies are as follows:

[0013] Wherein, A is the first sector, B is the second sector, C is the third sector, and D is the fourth sector; the echo amplitude and echo energy are taken as the original values ​​or as normalized changes relative to the reference state to correct for differences in initial transducer efficiency, channel gain, or coupling state in different sectors.

[0014] It should be noted that the corresponding echo phase or spectral features need to be further extracted as needed.

[0015] Furthermore, the composite feature construction module is used to construct the axial force sensitive mean feature. Since the stress on the bolt section is approximately uniform under axial load, the ToF changes in the four sectors are consistent. Therefore, the axial force sensitive mean feature can be expressed as:

[0016] in, Used to characterize changes in axial load or axial preload.

[0017] Furthermore, the composite feature construction module is used to construct moment-sensitive opposing differential features. Since the bolt section is under tension on one side and compression on the other side under bending moment, the ultrasonic flight time changes in opposing sectors are different. Therefore, the moment-sensitive features in the two orthogonal directions can be expressed as follows:

[0018]

[0019] in, , These are used to characterize the bending response in two orthogonal directions.

[0020] Furthermore, the composite feature construction module is used to construct shear-sensitive neighborhood asymmetric features. Under shear force, the bolt undergoes lateral deformation, interface micro-slippage, local contact state changes, and echo energy distribution asymmetry. The shear-sensitive features include one or more of the following: ToF neighborhood asymmetric features, amplitude asymmetric features, and energy asymmetric features. The asymmetric features of the ToF neighborhood related to the shear force direction are represented as follows:

[0021]

[0022] in, and This represents the circumferential ToF asymmetric response caused by shearing in different directions; The magnitude asymmetry characteristic related to shear force direction is represented as follows:

[0023]

[0024] in, and This indicates asymmetry in amplitude caused by shearing forces acting in different directions. The energy asymmetry characteristic related to the shear force direction is represented as follows:

[0025]

[0026] in, and This indicates the energy asymmetry caused by different shearing actions; The composite feature construction module combines the aforementioned axial force-sensitive mean feature, bending moment-sensitive opposing difference feature, shear force-sensitive neighborhood asymmetric feature, and amplitude / energy perturbation feature into a composite ultrasonic feature vector: .

[0027] Furthermore, the composite load decoupling module takes the composite ultrasound feature vector as input and outputs a composite load vector:

[0028] in, This is an axial load or axial preload. and These are the shear force components in two orthogonal directions. and These are the bending moment components in two orthogonal directions; The composite load decoupling model can be either a nonlinear or a linear model. When a nonlinear model is used, the following conditions must be met:

[0029] in, For multinomial regression models, lookup table models, support vector regression models, neural network models, finite element-assisted calibration models, or other data-driven models; When using a linear model:

[0030] in, For the composite load decoupling matrix, It is the bias vector; A local coordinate system is established with the direction of the intelligent bolt axis as the z-axis, the direction of the line connecting A and C as the x-axis, and the direction of the line connecting B and D as the y-axis. Under the local coordinate system, the first direction shear force, the second direction shear force, the first direction bending moment, and the second direction bending moment are defined.

[0031] Furthermore, a method for ultrasonic decoupling monitoring of composite loads on intelligent bolts is characterized by the following steps: Step S100: A four-sector thin-film piezoelectric ultrasonic transducer array is set on the head end face of the smart bolt, the array including a first sector, a second sector, a third sector and a fourth sector arranged circumferentially; Step S200: Acquire ultrasonic echo signals from four sectors using a multi-channel ultrasonic excitation and receiving module; Step S300: Extract the ultrasonic time-of-flight variation, echo amplitude, echo energy, phase or spectral characteristics from the ultrasonic echo signals of the four sectors respectively; Step S400: Construct axial force sensitive mean characteristics based on the ultrasonic time-of-flight variations of the four sectors; Step S500: Construct moment-sensitive opposing differential features based on the differences in ultrasonic flight time variations between opposing sectors; Step S600: Construct shear-sensitive neighborhood asymmetric features based on the changes in ultrasonic time-of-flight, echo amplitude, or echo energy differences between adjacent or neighboring sectors; Step S700: Combine the axial force sensitive mean feature, bending moment sensitive opposing difference feature, shear force sensitive neighborhood asymmetric feature, and amplitude / energy perturbation feature into a composite ultrasonic feature vector; Step S800: Input the composite ultrasound feature vector into the composite load decoupling model; Step S900: Output the axial force, shear force in two orthogonal directions, and bending moment in two orthogonal directions of the smart bolt.

[0032] Furthermore, before establishing the composite load decoupling model, the smart bolt is calibrated. The calibration process includes: applying multiple levels of axial load individually to obtain the relationship between the axial force sensitivity mean characteristics and the axial load; applying shear forces in two orthogonal directions to obtain the relationship between the shear force sensitivity neighborhood asymmetry characteristics, amplitude difference, energy difference, and shear force; applying bending moments in two orthogonal directions to obtain the relationship between the bending moment sensitivity opposing difference characteristics and the bending moment; applying axial force-shear force, axial force-bending moment, shear force-bending moment, and axial force-shear force-bending moment composite loading to obtain the cross-sensitivity relationship between load components, and establishing or modifying the composite load decoupling matrix accordingly.

[0033] Furthermore, the composite load decoupling model includes a cross-sensitivity correction mechanism; the cross-sensitivity coefficients of different load components to each ultrasonic feature are obtained through composite loading calibration, and the coupling effect between axial force, shear force and bending moment is corrected during load inversion to improve the decoupling accuracy under composite loading conditions.

[0034] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: This invention acquires multi-channel ultrasonic responses in different circumferential orientations using a four-sector thin-film piezoelectric ultrasonic transducer array, overcoming the limitation of a single ultrasonic channel only reflecting local or axial stress states. Furthermore, by constructing a four-channel Time-of-Flight (ToF) mean feature to characterize axial load, it helps to reduce the influence of local disturbances and bending differences on axial load estimation, improving the stability of axial force monitoring. By constructing opposing sector ToF differential features to characterize the bending moment response in two orthogonal directions, it can identify circumferentially non-uniform ultrasonic responses caused by tension on one side and compression on the other side of a bolt section under bending. Finally, by constructing asymmetric features in the shear-sensitive neighborhood and combining them with echo amplitude, energy, phase, or spectrum... Auxiliary features can improve the ability to identify transverse shear loads and avoid insufficient shear force identification due to relying solely on single acoustic time difference features. By using composite ultrasonic feature vectors to structurally express the ultrasonic responses related to axial force, shear force, and bending moment, a multi-dimensional input basis is provided for the decoupled inversion of composite loads. By synchronously outputting axial force, bidirectional shear force, and bidirectional bending moment through the composite load decoupling model, the actual loading state of the smart bolt can be reflected more comprehensively than single preload monitoring or single shear force measurement. By establishing a decoupling model through joint calibration of individual loads and composite loads, the cross-sensitive influence between axial force, shear force, and bending moment can be considered, improving the inversion accuracy under composite loading conditions. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the intelligent bolt composite load ultrasonic decoupling monitoring system of the present invention; Figure 2 This is a schematic diagram of the end face structure of the four-sector thin-film piezoelectric ultrasonic transducer array of the present invention; Figure 3 This is a schematic diagram of the multi-channel ultrasonic response feature extraction of the present invention; Figure 4 This is a schematic diagram illustrating the axial force, shear force, and bending moment sensitive features of the present invention. Figure 5 This is a schematic diagram of the composite load decoupling model and calibration process of the present invention.

[0036] The diagram is labeled as follows: 1. Smart bolt; 2. Four-sector thin-film piezoelectric ultrasonic transducer array; A, B, C, and D are four sector-shaped thin-film piezoelectric ultrasonic transducer regions arranged circumferentially along the head end face of the smart bolt, respectively; 2a, first sector thin-film piezoelectric ultrasonic transducer region; 2b, second sector thin-film piezoelectric ultrasonic transducer region; 2c, third sector thin-film piezoelectric ultrasonic transducer region; 2d, fourth sector thin-film piezoelectric ultrasonic transducer region. Detailed Implementation

[0037] The present invention will now be described in detail with reference to the accompanying drawings.

[0038] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0039] This invention obtains multi-channel ultrasonic responses in different circumferential orientations by using a four-sector thin-film piezoelectric ultrasonic transducer array set on the head end face of a smart bolt; extracts features such as ultrasonic time-of-flight variation, echo amplitude, echo energy, echo phase, or spectrum from the echoes of each channel; constructs axial force sensitive features using the four-channel mean, constructs bending moment sensitive features using the opposing difference, and constructs shear force sensitive features using neighborhood asymmetry and amplitude / energy differences; further, it forms a composite ultrasonic feature vector and inputs it into a composite load decoupling model, finally outputting axial force, bidirectional shear force, and bidirectional bending moment.

[0040] Example 1 like Figure 1 As shown, this embodiment provides an intelligent bolt composite load ultrasonic decoupling monitoring system. The system includes an intelligent bolt 1, a four-sector thin-film piezoelectric ultrasonic transducer array 2, a multi-channel ultrasonic excitation and receiving module, a multi-channel ultrasonic feature extraction module, a composite feature construction module, a composite load decoupling module, a calibration parameter storage module, and a load status output module.

[0041] The smart bolt 1 is the object to be monitored and can be a key connecting fastener in aerospace equipment, weaponry, wind turbines, hydroelectric turbines, nuclear power equipment, rail transit equipment, energy and transportation equipment, or other major engineering equipment. A four-sector thin-film piezoelectric ultrasonic transducer array 2 is disposed on the head end face of the smart bolt 1 to acquire ultrasonic responses in different circumferential orientations. A multi-channel ultrasonic excitation and reception module is used to control the four-sector thin-film piezoelectric ultrasonic transducer array 2 to perform independent excitation, independent reception, or time-division multiplexing transmission and reception, and to obtain ultrasonic echo signals from multiple channels.

[0042] like Figure 1 As shown, the four-sector thin-film piezoelectric ultrasonic transducer array 2 is connected to the multi-channel ultrasonic excitation and receiving module. The multi-channel ultrasonic excitation and receiving module can be configured according to a preset channel order, polling mode, or time-division multiplexing mode, such as... Figure 2 As shown, ultrasonic excitation and echo reception are performed on the first sector thin-film piezoelectric ultrasonic transducer region 2a, the second sector thin-film piezoelectric ultrasonic transducer region 2b, the third sector thin-film piezoelectric ultrasonic transducer region 2c, and the fourth sector thin-film piezoelectric ultrasonic transducer region 2d. The multi-channel ultrasonic signal output from the four-sector thin-film piezoelectric ultrasonic transducer array 2 enters the multi-channel ultrasonic excitation and reception module, and is sequentially transmitted to the multi-channel ultrasonic feature extraction module, the composite feature construction module, the composite load decoupling module, and the load status output module.

[0043] The multi-channel ultrasound feature extraction module processes the ultrasound echo signals from each channel to extract the ultrasound time-of-flight variation, echo amplitude, echo energy, and echo phase or spectral features. The composite feature construction module constructs axial force-sensitive mean features, bending moment-sensitive opposing difference features, shear force-sensitive neighborhood asymmetric features, and amplitude / energy perturbation features based on the multi-channel ultrasound features of each sector, and forms a composite ultrasound feature vector.

[0044] The composite load decoupling module outputs axial force, shear force in the first direction, shear force in the second direction, bending moment in the first direction, and bending moment in the second direction based on the composite ultrasonic eigenvector and the composite load decoupling model. The calibration parameter storage module stores the composite load decoupling matrix, bias vector, cross sensitivity coefficient, or model parameters. The load state output module outputs the composite load state of smart bolt 1.

[0045] Through the modules described above, this embodiment forms a complete technical chain from four-sector multi-channel ultrasonic acquisition, multi-channel feature extraction, load-sensitive feature construction, composite ultrasonic feature vector formation to composite load decoupling output. The focus of this system is not on protecting the four-sector structure itself, nor on measuring a single type of load using a single acoustic time-of-flight measurement, but rather on utilizing the mean, opposing differences, and neighborhood asymmetry relationships between the ultrasonic responses of the four sectors to achieve synchronous decoupling monitoring of axial force, shear force, and bending moment.

[0046] Example 2 like Figure 2 As shown, a four-sector thin-film piezoelectric ultrasonic transducer array 2 is disposed on the head end face of the smart bolt 1. The array includes a first sector thin-film piezoelectric ultrasonic transducer region 2a, a second sector thin-film piezoelectric ultrasonic transducer region 2b, a third sector thin-film piezoelectric ultrasonic transducer region 2c, and a fourth sector thin-film piezoelectric ultrasonic transducer region 2d.

[0047] In a preferred embodiment, four sector-shaped thin-film piezoelectric ultrasonic transducer regions are evenly arranged along the circumferential direction of the head end face of the smart bolt 1, corresponding to the four directions A, B, C, and D respectively. The first sector thin-film piezoelectric ultrasonic transducer region 2a corresponds to channel A, the second sector thin-film piezoelectric ultrasonic transducer region 2b corresponds to channel B, the third sector thin-film piezoelectric ultrasonic transducer region 2c corresponds to channel C, and the fourth sector thin-film piezoelectric ultrasonic transducer region 2d corresponds to channel D. Channels A and C form one pair of opposing sectors, and channels B and D form another pair of opposing sectors. For ease of description, a local coordinate system is established with the axis of the smart bolt 1 as the z-axis, the line connecting channels A and C as the x-axis, and the line connecting channels B and D as the y-axis. and These represent the shear force components in two orthogonal directions, i.e., the shear force in the first direction. Second direction shear force , and These represent the bending moment components in two orthogonal directions, with the bending moment in the first direction being... Second direction bending moment The coordinate system described above is only used to illustrate the feature structure and composite load decoupling relationship of this embodiment. In practical applications, coordinate transformation can be performed according to the installation direction of the smart bolt and the equipment coordinate system. The relative sector relationship described above is used to construct moment-sensitive opposing differential features in the subsequent construction, and the circumferential neighborhood relationship is used to construct shear-sensitive neighborhood asymmetric features in the subsequent construction.

[0048] Four thin-film piezoelectric ultrasonic transducer regions can be integrated onto the head end face of the smart bolt 1 using physical vapor deposition, magnetron sputtering, ion plating, or other thin-film fabrication processes. Each sector can serve as an independent ultrasonic excitation and echo receiving region, or a time-division multiplexing method can be used to achieve integrated transmission and reception within the same sector. Insulation gaps can be set between the four sectors to reduce electrical coupling interference and signal crosstalk between different channels.

[0049] In practical implementation, the four-sector thin-film piezoelectric ultrasonic transducer array 2 can be arranged around the central region of the head end face of the smart bolt, enabling the ultrasonic waves corresponding to each sector to propagate along the bolt axis and return, while preserving the sensitivity differences of different circumferential orientations to local stress, bending, and shear states. The uniform arrangement of the four sectors is beneficial for constructing four-channel mean characteristics, two sets of opposing differential characteristics, and circumferential neighborhood asymmetric characteristics, thus providing a structural basis for decoupling composite loads.

[0050] It should be noted that this embodiment does not necessarily limit the specific thin film material, thin film thickness, or preparation process. The four-sector thin-film piezoelectric ultrasonic transducer array 2 can serve as the basis for this embodiment as long as it can acquire distinguishable ultrasonic responses in different circumferential orientations. In other embodiments, the number of sectors can also be an even number greater than four. When the number of sectors is greater than four, relative sector bending moment sensitive features can be selected, adjacent sector or neighboring sector shear force sensitive features can be selected, and composite load decoupling can be performed through multi-channel feature vectors.

[0051] Example 3 like Figure 3 As shown, the multi-channel ultrasonic excitation and receiving module controls the channels of the four-sector thin-film piezoelectric ultrasonic transducer array 2 to obtain ultrasonic echo signals from channels A, B, C, and D, respectively. The multi-channel ultrasonic feature extraction module processes the ultrasonic echo signals from the four channels and outputs the ultrasonic time-of-flight variation, echo amplitude, echo energy, and echo phase or spectral characteristics of each channel.

[0052] Let the changes in ultrasonic time-of-flight for the four sectors A, B, C, and D be as follows:

[0053] The corresponding echo amplitudes are as follows:

[0054] The corresponding echo energies are as follows:

[0055] Where A is the first sector, B is the second sector, C is the third sector, and D is the fourth sector.

[0056] Among them, the change in ultrasonic flight time can be obtained by the change in the arrival time of the target echo in each channel relative to the reference state; the echo amplitude can be the main peak amplitude, envelope peak value, or peak-to-peak value of the target echo; the echo energy can be the sum of squares of the signal within the target echo window, the envelope integral, or the energy of a specific frequency band; the echo phase or spectral characteristics can include phase difference, center frequency, main frequency amplitude, frequency band energy, main peak width, spectral shift, or mode conversion wave energy, etc.

[0057] To mitigate the impact of differences in thin-film transducer efficiency, channel gain, and initial coupling state across different sectors on amplitude and energy characteristics, echo amplitude and echo energy can be expressed as their original values ​​or as normalized variations relative to the reference state. Let the echo amplitude and echo energy of the i-th channel under the reference state be respectively... and The echo amplitude and echo energy in real time are respectively and ,in, Then it can be adopted As an amplitude normalization feature, adopt As an energy normalization feature, subsequent amplitude asymmetry features and energy asymmetry features can be constructed based on the original amplitude and original energy, or based on the normalized amplitude change and energy change.

[0058] In practical implementation, the ultrasonic echo signals of channels A, B, C, and D can be acquired under the same reference state and then acquired in real time under different load states. The multi-channel ultrasonic feature extraction module can calculate the ToF change, amplitude change, energy change, or phase change of each channel relative to the reference state, and use these as inputs to the composite feature construction module.

[0059] It should be noted that this embodiment does not limit the specific ToF extraction algorithm. The ToF changes of each channel can be obtained using existing time delay estimation methods, or using target echo window truncation, cross-correlation, peak localization, phase estimation, or other high-precision ToF extraction methods. The focus of this embodiment is to construct composite load-sensitive features using the ultrasonic response characteristics such as ToF, amplitude, energy, and phase of the four channels, and to invert axial force, shear force, and bending moment through a decoupled model.

[0060] Example 4 like Figure 4 As shown, the composite feature construction module constructs axial force sensitive mean features, bending moment sensitive opposing difference features, shear force sensitive neighborhood asymmetric features, and amplitude / energy perturbation features based on the multi-channel ultrasonic response of the four sectors.

[0061] First, the axial force sensitivity mean characteristic is constructed. Under axial load, the stress distribution on the bolt section is approximately uniform, and the Time-of-F (ToF) variations in the four channels A, B, C, and D are consistent. Therefore, the axial force sensitivity mean characteristic can be expressed as:

[0062] in, It is a mean characteristic sensitive to axial force, used to characterize changes in axial load or axial preload.

[0063] Compared to single-channel ToF features, four-channel mean features can reduce the impact of local disturbances or single-channel anomalies on axial force estimation.

[0064] Secondly, a moment-sensitive opposing differential characteristic is constructed. Under bending moment, the bolt section is under tension on one side and compression on the other, resulting in differences in the ToF changes of the opposing sectors. Therefore, the moment-sensitive opposing differential characteristic can be expressed as:

[0065]

[0066] in, , These are used to characterize the bending response in two orthogonal directions; Channel A and Channel C form a set of opposing sectors to reflect the difference in opposing response caused by bending in the first direction; Channel B and Channel D form another set of opposing sectors to reflect the difference in opposing response caused by bending in the second direction.

[0067] Secondly, we can construct asymmetric features in the shear-sensitive neighborhood. Under shear force, bolts may exhibit lateral deformation, changes in local contact state, micro-slip at the interface, and asymmetric energy distribution. Therefore, we can construct asymmetric features in the ToF neighborhood.

[0068]

[0069] in, and This represents the circumferential ToF asymmetric response caused by shearing in different directions.

[0070] Shear force can cause not only changes in Time of Flight (ToF), but also attenuation of echo amplitude and changes in energy distribution. Therefore, amplitude asymmetry characteristics can be constructed:

[0071]

[0072] The energy asymmetry characteristic related to the shear force direction is represented as follows:

[0073]

[0074] The aforementioned amplitude asymmetry and energy asymmetry characteristics together constitute the amplitude / energy perturbation characteristics. These characteristics are mainly used to characterize changes in contact state, local energy attenuation, changes in echo propagation paths, or mode conversion effects caused by lateral shear force.

[0075] It should be noted that the aforementioned shear-sensitive neighborhood asymmetric features, amplitude asymmetric features, and energy asymmetric features are only preferred construction methods. In practical applications, other adjacent sector combinations or weighted combinations can be selected based on the spatial orientation of the four sectors, the definition of the shear direction, calibration results, and channel sensitivity. For example, weighting coefficients can be set for different channels before constructing asymmetric features to correct for differences in the initial sensitivity of each sector. As long as the feature can reflect the circumferential asymmetric ultrasonic response caused by transverse shear force, it can be used as the shear-sensitive feature of this invention. The shear-sensitive feature can be used alone or in combination with amplitude, energy, phase, or spectral features.

[0076] like Figure 4 As shown, the composite feature construction module does not directly input the original ultrasonic features of the four channels into the decoupling model, but first converts them into structured features with explicit load sensitivity. Specifically, the axial force-sensitive mean feature mainly characterizes the common changes of the four channels; the bending moment-sensitive opposing difference feature mainly characterizes the tensile and compressive differences between relative sectors; and the shear force-sensitive neighborhood asymmetric feature and amplitude / energy perturbation feature mainly characterize the circumferential asymmetric response caused by lateral shear force. This feature construction process reduces the problem of single-channel features being sensitive to the cross-sensitivity of multiple load components, providing a stable input for subsequent composite load decoupling.

[0077] Example 5 like Figure 5As shown, the composite feature construction module combines axial force-sensitive mean features, bending moment-sensitive opposing difference features, shear force-sensitive neighborhood asymmetric features, and amplitude / energy perturbation features to form a composite ultrasonic feature vector. As an optional implementation, the composite ultrasonic feature vector can be expressed as:

[0078] In other embodiments, the composite ultrasound feature vector may also include phase difference, spectral center frequency, main peak width, mode conversion wave energy, echo attenuation coefficient, multi-channel consistency index, or other ultrasound features that can reflect the composite load state.

[0079] The composite load decoupling module inputs the composite ultrasound feature vector into the composite load decoupling model and outputs the composite load vector:

[0080] in, This is an axial load or axial preload. and These are the shear force components in two orthogonal directions. and These are the bending moments in the first and second directions, respectively.

[0081] like Figure 5 As shown, the input to the composite load decoupling model is a composite ultrasonic feature vector, and the output includes axial force output, shear force output in the first direction, shear force output in the second direction, bending moment output in the first direction, and bending moment output in the second direction. The composite loading calibration module is used to obtain the decoupling matrix or model parameters based on the axial loading calibration, shear loading calibration, bending moment loading calibration, and composite loading calibration data. The decoupling matrix or model parameters are input into the composite load decoupling model to achieve synchronous inversion of axial force, bidirectional shear force, and bidirectional bending moment.

[0082] As an alternative implementation, the composite load decoupling model can adopt a linear decoupling model:

[0083] in, For the composite load decoupling matrix, The bias vector. Composite load decoupling matrix. and bias vector It can be obtained through calibration and stored in the calibration parameter storage module.

[0084] As another alternative implementation, the composite load decoupling model can adopt a nonlinear model:

[0085] in, This includes multinomial regression models, lookup table models, support vector regression models, neural network models, finite element-assisted calibration models, or other data-driven models.

[0086] In obtaining and Then, the resultant bending moment and the direction of the bending moment can be further calculated:

[0087]

[0088] In obtaining and Then, the resultant shear force and the direction of the shear force can be further calculated:

[0089]

[0090] The axial force, shear force component, shear force resultant, shear force direction, bending moment component, bending moment resultant, and bending moment direction mentioned above can all be used as output content of the load state output module.

[0091] Example 6 like Figure 5 As shown, the composite loading calibration module is used to obtain the decoupling matrix or model parameters required for the composite load decoupling model. The composite loading calibration module receives data from axial loading calibration, shear loading calibration, bending moment loading calibration, and composite loading calibration, and calculates or trains the decoupling matrix or model parameters based on the above data.

[0092] Before actual use, a composite load decoupling model can be established using a calibration device. The calibration device may include an axial loading device, a transverse shear loading device, a bending moment loading device, and a multi-axis composite loading device.

[0093] The calibration process may include the following steps: First, axial loading calibration. Multiple levels of axial load or axial preload are applied to smart bolt 1, and the ultrasonic response in four sectors is collected and extracted. And calculate the axial force sensitivity mean characteristics. This allows us to obtain the correspondence between the axial force sensitivity mean characteristics and the axial load.

[0094] Second, bending moment loading calibration. Multiple levels of bending moment were applied to the smart bolt 1 along two orthogonal directions, and the ultrasonic response in four sectors was collected and calculated. and The relationship between moment-sensitive opposing differential features and corresponding bending moments is obtained.

[0095] Third, shear loading calibration. Multiple levels of shear force were applied to the smart bolt 1 along two orthogonal directions, and the ultrasonic response in four sectors was collected and calculated. , , , , , By studying these characteristics, the relationship between the asymmetric features of the shear-sensitive neighborhood and the amplitude / energy perturbation features and the shear force in the corresponding direction can be obtained.

[0096] Fourth, composite loading calibration. Axial force-shear force, axial force-bending moment, shear force-bending moment, and composite loading of axial force-shear force-bending moment are applied to obtain the multi-channel ultrasonic response and composite ultrasonic characteristic vector under composite loading conditions.

[0097] Fifth, decoupling model establishment. Based on the individual load calibration and composite load calibration data mentioned above, a composite load decoupling matrix is ​​established. Bias vector or nonlinear decoupling model It generates a decoupling matrix or model parameters, which are stored in the calibration parameter storage module or directly input into the composite load decoupling model.

[0098] Through the above calibration process, the response relationship and cross-sensitivity relationship of different load components to each ultrasonic feature can be obtained, thereby correcting the coupling effect between axial force, shear force and bending moment during the load inversion process.

[0099] Example 7 Under actual combined loading conditions, axial force, shear force, and bending moment do not have completely independent effects on the ultrasonic response of each channel. Axial force may simultaneously affect the mean Time-of-Flight (ToF) and local energy levels of all four sectors; bending moment may affect the differential signal in opposing sectors and may also alter neighborhood asymmetry characteristics; shear force may cause combined changes in ToF, amplitude, energy, and phase. Therefore, this invention can further introduce a cross-sensitivity correction mechanism.

[0100] As an optional implementation, the composite load decoupling matrix may include coefficients describing the cross-influence between different load components. The composite load decoupling matrix can be represented as a matrix structure consisting of multiple submatrices:

[0101] in, Used to characterize the contribution of axial force features to axial force inversion. Used to characterize the contribution of shear force features to shear force inversion. Used to characterize the contribution of bending moment features to bending moment inversion; , , , , , Used to characterize the cross-sensitivity relationship between different load components.

[0102] After obtaining the aforementioned cross-sensitivity coefficients through composite loading calibration, the composite load decoupling module can correct the coupling effects between load components during the inversion process, thereby improving the composite load decoupling accuracy. This cross-sensitivity correction mechanism can avoid misjudging the common changes in the four channels caused by axial force as bending moment or shear force, and can also reduce the cross-interference between bending moment and shear force sensitivity characteristics, and between shear force and bending moment sensitivity characteristics.

[0103] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent bolt composite load ultrasonic decoupling monitoring system, characterized in that, The system includes the following: Smart bolts are used as monitored connection fasteners. A four-sector thin-film piezoelectric ultrasonic transducer array is disposed on the end face of the head of a smart bolt, including a first sector, a second sector, a third sector, and a fourth sector arranged circumferentially, for obtaining ultrasonic response signals in different circumferential orientations respectively. The multi-channel ultrasonic excitation and reception module is used to independently excite, independently receive, or time-division multiplex the four-sector thin-film piezoelectric ultrasonic transducer region and obtain ultrasonic echo signals from four channels. The multi-channel ultrasound feature extraction module is used to extract ultrasound time-of-flight variation, echo amplitude, echo energy, phase, spectral features, or mode conversion features from ultrasound echo signals from each channel. The composite feature construction module is used to construct axial force-sensitive mean features, bending moment-sensitive opposing difference features, shear force-sensitive neighborhood asymmetric features, and amplitude / energy perturbation features based on the multi-channel ultrasonic features of four sectors, and form a composite ultrasonic feature vector. The composite load decoupling module is used to input the composite ultrasonic feature vector into the composite load decoupling model and output the axial force, bidirectional shear force, and bidirectional bending moment experienced by the smart bolt. The bidirectional shear force includes shear force in a first direction and shear force in a second direction, and the bidirectional bending moment includes bending moment in a first direction and bending moment in a second direction. The calibration parameter storage module is used to store composite load decoupling matrices, bias vectors, cross sensitivity coefficients, or model parameters. The load status output module is used to output the composite load status of the smart bolt, including at least one of axial force, shear force in the first direction, shear force in the second direction, bending moment in the first direction, and bending moment in the second direction.

2. The intelligent bolt composite load ultrasonic decoupling monitoring system according to claim 1, characterized in that, The four-sector thin-film piezoelectric ultrasonic transducer array is set in the central region of the hexagonal head end face of the smart bolt or in the annular region surrounding the central region of the end face; the four sector-shaped thin-film piezoelectric ultrasonic transducer regions are evenly arranged along the circumference and are denoted as A, B, C, and D, respectively, corresponding to 0°, 90°, 180°, and 270° azimuths; the first sector A and the third sector C form a pair of opposing sectors, and the second sector B and the fourth sector D form another pair of opposing sectors.

3. The intelligent bolt composite load ultrasonic decoupling monitoring system according to claim 2, characterized in that, it is set that... The changes in ultrasonic time-of-flight for the four sectors are as follows: The corresponding echo amplitudes are as follows: The corresponding echo energies are as follows: Wherein, A is the first sector, B is the second sector, C is the third sector, and D is the fourth sector; the echo amplitude and echo energy are taken as the original values ​​or as normalized changes relative to the reference state to correct for differences in initial transducer efficiency, channel gain, or coupling state in different sectors.

4. The intelligent bolt composite load ultrasonic decoupling monitoring system according to claim 3, characterized in that, The composite feature construction module is used to construct the axial force sensitive mean feature. Since the stress on the bolt section is approximately uniform under axial load, the ToF changes in the four sectors are consistent. Therefore, the axial force sensitive mean feature can be expressed as: in, Used to characterize changes in axial load or axial preload.

5. The intelligent bolt composite load ultrasonic decoupling monitoring system according to claim 4, characterized in that, The composite feature construction module is used to construct moment-sensitive opposing differential features. Since the bolt section is under tension on one side and compression on the other under bending moment, the ultrasonic flight times of opposing sectors differ. Therefore, the moment-sensitive features in the two orthogonal directions can be expressed as follows: in, , These are used to characterize the bending response in two orthogonal directions.

6. The intelligent bolt composite load ultrasonic decoupling monitoring system according to claim 5, characterized in that, The composite feature construction module is used to construct shear-sensitive neighborhood asymmetric features. Under shear force, bolts exhibit lateral deformation, interface micro-slippage, local contact state changes, and asymmetric echo energy distribution. Shear-sensitive features include one or more of the following: ToF neighborhood asymmetric features, amplitude asymmetric features, and energy asymmetric features. The asymmetric features of the ToF neighborhood related to the shear force direction are represented as follows: in, and This represents the circumferential ToF asymmetric response caused by shearing in different directions; The magnitude asymmetry characteristic related to shear force direction is represented as follows: in, and This indicates asymmetry in amplitude caused by shearing forces acting in different directions. The energy asymmetry characteristic related to the shear force direction is represented as follows: in, and This indicates the energy asymmetry caused by different shearing actions; The composite feature construction module combines the aforementioned axial force-sensitive mean feature, bending moment-sensitive opposing difference feature, shear force-sensitive neighborhood asymmetric feature, and amplitude / energy perturbation feature into a composite ultrasonic feature vector: 。 7. The intelligent bolt composite load ultrasonic decoupling monitoring system according to claim 6, characterized in that, The composite load decoupling module takes the composite ultrasound feature vector as input and outputs the composite load vector: in, This is an axial load or axial preload. and These are the shear force components in two orthogonal directions. and These are the bending moment components in two orthogonal directions; The composite load decoupling model can be either a nonlinear or a linear model. When a nonlinear model is used, the following conditions must be met: in, For multinomial regression models, lookup table models, support vector regression models, neural network models, finite element-assisted calibration models, or other data-driven models; When using a linear model: in, For the composite load decoupling matrix, It is the bias vector; A local coordinate system is established with the direction of the intelligent bolt axis as the z-axis, the direction of the line connecting A and C as the x-axis, and the direction of the line connecting B and D as the y-axis. Under the local coordinate system, the first direction shear force, the second direction shear force, the first direction bending moment, and the second direction bending moment are defined.

8. A method for ultrasonic decoupling monitoring of intelligent bolt composite loads, employing the intelligent bolt composite load ultrasonic decoupling monitoring system according to any one of claims 1-7, characterized in that, The method includes the following steps: Step S100: A four-sector thin-film piezoelectric ultrasonic transducer array is set on the head end face of the smart bolt, the array including a first sector, a second sector, a third sector and a fourth sector arranged circumferentially; Step S200: Acquire ultrasonic echo signals from four sectors using a multi-channel ultrasonic excitation and receiving module; Step S300: Extract the ultrasonic time-of-flight variation, echo amplitude, echo energy, phase or spectral characteristics from the ultrasonic echo signals of the four sectors respectively; Step S400: Construct axial force sensitive mean characteristics based on the ultrasonic time-of-flight variations of the four sectors; Step S500: Construct moment-sensitive opposing differential features based on the differences in ultrasonic flight time variations between opposing sectors; Step S600: Construct shear-sensitive neighborhood asymmetric features based on the changes in ultrasonic time-of-flight, echo amplitude, or echo energy differences between adjacent or neighboring sectors; Step S700: Combine the axial force sensitive mean feature, bending moment sensitive opposing difference feature, shear force sensitive neighborhood asymmetric feature, and amplitude / energy perturbation feature into a composite ultrasonic feature vector; Step S800: Input the composite ultrasound feature vector into the composite load decoupling model; Step S900: Output the axial force, shear force in two orthogonal directions, and bending moment in two orthogonal directions of the smart bolt.

9. The method for ultrasonic decoupling monitoring of intelligent bolt composite loads according to claim 8, characterized in that, Before establishing the composite load decoupling model, the smart bolt is calibrated. The calibration process includes: applying multiple levels of axial load individually to obtain the relationship between the axial force sensitivity mean characteristics and the axial load; applying shear forces in two orthogonal directions to obtain the relationship between the shear force sensitivity neighborhood asymmetry characteristics, amplitude difference, energy difference and shear force; applying bending moments in two orthogonal directions to obtain the relationship between the bending moment sensitivity opposing difference characteristics and the bending moment; applying axial force-shear force, axial force-bending moment, shear force-bending moment and axial force-shear force-bending moment composite loading to obtain the cross-sensitivity relationship between load components, and establishing or modifying the composite load decoupling matrix accordingly.

10. The method for ultrasonic decoupling monitoring of intelligent bolt composite loads according to claim 9, characterized in that, The composite load decoupling model includes a cross-sensitivity correction mechanism; the cross-sensitivity coefficients of different load components to each ultrasonic feature are obtained through composite loading calibration, and the coupling effect between axial force, shear force and bending moment is corrected during load inversion to improve the decoupling accuracy under composite loading conditions.

Images