Ultrasonic axial force high-speed real-time measurement system

Abstract

The invention relates to the technical field of ultrasonic stress measurement, and discloses an ultrasonic axial force high-speed real-time measurement system which comprises a core processor module which controls a transducer to alternately transmit and receive longitudinal wave and transverse wave echo signals and calls a fingerprint template established based on a geometric structure of a bolt to perform cross-correlation operation so as to determine transit time, according to the method, by constructing the acoustic characteristic normalization ratio, the influence of uniform temperature change is inhibited on the aspect of the measurement principle, and the inherent problems of model misalignment and error amplification in a traditional compensation mode are avoided; echo positioning is carried out by using geometric fingerprints of the bolt, a filtering algorithm with high calculation amount is replaced, the reliability of the system is improved, and high-precision measurement can be realized on a low-power-consumption general processor.

Description

TECHNICAL FIELD

[0001] The present application relates to an ultrasonic axial force high-speed real-time measurement system, and belongs to the technical field of ultrasonic stress measurement. BACKGROUND

[0002] Currently, in the aspect of ensuring the safety of large structures and key equipment, using ultrasonic technology to measure the axial pretightening force of bolts with high precision and non-destructively is an important technical method. Currently, the commonly used measurement logic is to first measure the propagation time of ultrasonic waves in the bolt, then obtain the temperature reading through an independent temperature sensor, and then compensate according to a mathematical model to invert the axial force.

[0003] However, the effectiveness of this measurement method highly depends on the ideal premise that the measured single-point temperature can accurately represent the overall temperature field of the bolt. In real engineering scenarios, when the bolt forms a non-uniform temperature gradient inside due to sunlight or adjacent heat sources, the compensation model based on surface temperature measurement will lose its physical basis, and the incorrect compensation value output will become the main error source of the measurement result, causing the axial force reading given by the system to deviate from the actual value and form a failure risk.

[0004] To address this problem, increasing the number of temperature sensors not only increases the complexity and cost of the system, but also cannot be implemented in many situations where drilling and modification of key bolts are prohibited, and still cannot accurately reflect the complex temperature field inside the bolt. Specifically, the existing technology mainly has the following deficiencies: 1. The measurement principle relies on the assumption of uniformity of the environment temperature, and in the temperature gradient working condition, the physical basis of the compensation model is invalidated, resulting in systematic measurement errors; 2. In the series logic of measurement and compensation, the small errors of sound time measurement and temperature measurement are all conducted and amplified in the compensation formula, which requires high precision of each link of the system, forming a contradiction between cost and precision; 3. To overcome noise interference, traditional echo signal processing usually relies on complex filtering algorithms, which increases the computational burden of the core processor and limits the application of the system in low-power long-endurance scenarios. Therefore, how to construct a measurement system whose generation process of measurement results itself has an inhibitory effect on temperature changes, and how to use a computationally efficient signal processing method to achieve low-cost and low-power real-time monitoring of axial force while ensuring high reliability, have become the technical problems to be solved by the present application. SUMMARY

[0005] The present application provides an ultrasonic axial force high-speed real-time measurement system, which mainly aims to solve the problem that the existing ultrasonic axial force measurement method relies on external temperature compensation and has a principle defect, and the signal processing computational burden is heavy, making it difficult to balance the measurement reliability, cost and power consumption.

[0006] To achieve the above objectives, the present invention provides a high-speed real-time ultrasonic axial force measurement system, comprising:

[0007] An electromagnetic ultrasonic transducer module is configured to alternately excite longitudinal wave ultrasonic pulses and transverse wave ultrasonic pulses in a bolt under test.

[0008] A core processor module is connected to an electromagnetic ultrasonic transducer module. The core processor module is configured to: control the electromagnetic ultrasonic transducer module to alternately transmit longitudinal and transverse ultrasonic pulses to the bolt under test; receive longitudinal and transverse echo signals propagated and returned by the bolt under test via the electromagnetic ultrasonic transducer module; invoke a geometric fingerprint template established and stored through a calibration process, the geometric fingerprint template being based on the fixed reflection characteristics formed by the geometric structure of the bolt under test, and perform cross-correlation operations between the received longitudinal and transverse echo signals and the geometric fingerprint template to determine the transit times of the longitudinal and transverse waves; calculate an acoustic feature normalization ratio based on the determined longitudinal and transverse wave transit times; and determine the axial force value of the bolt under test based on the mapping relationship between the acoustic feature normalization ratio established through the calibration process and the bolt axial force.

[0009] Preferably, the core processor module is configured to calculate the acoustic feature normalization ratio R in the following manner: Among them, v L For the longitudinal wave speed of sound, v S For the speed of sound of transverse waves, t L Let t be the transit time of the longitudinal wave. S This is the transit time of the transverse wave.

[0010] Preferably, the core processor module is configured to extract a fixed reflected echo waveform with a stable shape generated by the thread and chamfer of the bolt under test after the bolt under test reaches the target axial force through the initial tightening operation, and store the waveform as a geometric fingerprint template; the cross-correlation operation uses the fixed reflected echo waveform as a reference to lock the position of the same source echo in the subsequently received echo signal.

[0011] Preferably, the core processor module is further configured to: acquire a residual signal generated by cross-correlation operation; calculate at least one parameter value selected from the group consisting of the energy value and information entropy value of the residual signal; compare the parameter value with a health residual benchmark value measured when the bolt under test is in an initial tightened state; and output an alarm indicating an abnormal connection status when the parameter value exceeds the health residual benchmark value.

[0012] Preferably, the core processor module is further configured to: obtain a mismatch value after cross-correlation between the quantized echo signal and the geometric fingerprint template; derive a compensation coefficient characterizing the axial temperature gradient of the bolt under test based on a defined functional relationship between the mismatch value and the axial temperature gradient; and use the compensation coefficient to correct the mapping relationship to compensate for the influence of the axial temperature gradient on the determination of the axial force value.

[0013] Preferably, the core processor module is further configured to: filter any received echo signal to obtain the amplitude of its fundamental frequency component and the amplitude of its second harmonic component in parallel; calculate a relative nonlinear factor based on the amplitude of the fundamental frequency component and the amplitude of the second harmonic component; monitor the changing trend of the relative nonlinear factor while the axial force value remains within a preset fluctuation range; and output an early warning indicating early material fatigue damage when the changing trend shows a continuous increase and exceeds the damage judgment threshold.

[0014] Preferably, the core processor module is further configured to: additionally monitor the drive current or voltage amplitude of the excitation coil of the electromagnetic ultrasonic transducer module; and use this electrical parameter to normalize the relative nonlinear factor in order to eliminate the interference of excitation energy fluctuations on the early warning.

[0015] Preferably, the core processor module is further configured to: segment the entire echo waveform of the second harmonic component along the time axis, with each segment corresponding to multiple segments of different lengths of the bolt to be tested; calculate the average growth slope of the second harmonic component amplitude within each segment; and identify the segment with the largest average growth slope as the high-risk area where material fatigue damage is most concentrated.

[0016] Preferably, the core processor module is a microcontroller with a built-in digital signal processing instruction set. The microcontroller is configured to control the electromagnetic ultrasonic transducer module to complete the time-division alternating transmission of longitudinal wave ultrasonic pulses and transverse wave ultrasonic pulses with millisecond-level timing intervals.

[0017] Preferably, the core processor module is further configured to: when the bolt under test is in its initial tightened state, control the electromagnetic ultrasonic transducer module to emit a series of probe acoustic packets and record the returned composite echo signal containing multiple reflections from the bolt under test and the connection structure to establish a healthy acoustic fingerprint; in subsequent operation, periodically emit the same probe acoustic packets and collect real-time echoes; by comparing the real-time echoes with the healthy acoustic fingerprint using a dynamic time warping algorithm, when the calculated waveform similarity is lower than a preset similarity threshold, also output an alarm indicating an abnormal connection status.

[0018] Compared with the prior art, the beneficial effects of the present invention are:

[0019] 1. This technical solution uses the complex geometric reflection based on the inherent invariance of the bolt itself as the full waveform matching template for cross-correlation calculation. This procedure, while achieving echo localization, also constructs a measurement benchmark with the ability to suppress long-term drift at a deeper level. In long-term monitoring applications, traditional methods often rely on single-point features such as signal thresholds or peak values ​​for localization. This type of measurement benchmark is easily affected by signal amplitude and shape drift caused by transducer aging, acoustic coupling attenuation, and even changes in system electrical characteristics, resulting in measurement artifacts unrelated to the true state of the measured structure, seriously affecting the reliability of long-term monitoring data. In contrast, this solution uses a complete waveform rich in detail and derived from the inherent geometry of the bolt as the cross-correlation template. To match the reference, the essence of its cross-correlation operation is to seek the best overlap of the overall waveform shape, rather than to locate a single amplitude point. Therefore, when the signal undergoes overall amplitude attenuation or slight shape changes due to long-term service, its positioning result based on the entire waveform shape remains highly stable. This technical approach transforms the measurement reference from a volatile signal characteristic into a stable reference derived from the physical structure of the object under test. This effectively suppresses errors introduced by the system's own aging drift at the signal processing level, ensuring that any change in the system's output can reflect the true change in the physical state of the measured bolt with high confidence. This provides a reliable data foundation for realizing structural health prediction and life assessment based on minute change trends.

[0020] 2. This invention constructs a novel method for axial force measurement. It no longer relies on external measurement and mathematical compensation of interfering quantities such as temperature. Instead, it simultaneously excites longitudinal and transverse waves in the bolt under test and calculates the normalized ratio of their acoustic characteristics. Utilizing the physical characteristic that the two waves have similar trends in their influence on temperature but vastly different effects on stress, the influence of uniform temperature changes is suppressed at the measurement principle level. Simultaneously, the system uses a fingerprint template based on the bolt's own geometry for cross-correlation calculations, replacing traditional filtering and noise reduction to determine the acoustic wave transit time. This avoids the computational burden of complex filtering algorithms, allowing the system to run on low-power general-purpose processors. Furthermore, it transforms the inherent geometric reflection characteristics of the bolt, previously considered noise, into effective information for accurately locating echoes, demonstrating the engineering feasibility and inherent self-consistency of the measurement logic.

[0021] 3. The measurement system established in this invention not only outputs axial force values ​​but also possesses the ability to conduct multi-dimensional investigations into its own operating status and the health status of the measured object. By analyzing the residual signals generated during cross-correlation calculations, the system can identify unexpected reflections introduced by abnormal bolt connection conditions, such as loosening. Thus, without adding any sensing hardware, it achieves monitoring of the integrity of the connection structure, improving the system's ability to distinguish between two different failure modes: reduced axial force and loose connection. This provides a basis for equipment maintenance decisions. Furthermore, by using the same set of collected echo signals and processing them in parallel, it can further decode... By extracting the nonlinear acoustic information of the materials contained within, the system expands its functionality from macroscopic stress state monitoring to microscopic material damage early warning. By calculating the relative relationship between the second harmonic component and the fundamental frequency component in the echo signal, the system obtains a nonlinear factor that can characterize early fatigue damage. Furthermore, by analyzing the morphological changes of the echo waveform of this second harmonic component along the time axis, the spatial distribution of damage along the bolt length can also be located. This integrated perception of linear and nonlinear information enables the system to diagnose the health status of bolts in addition to its measurement function, extending its monitoring dimension from the current state to future trends.

[0022] 4. This invention uses an electromagnetic ultrasonic transducer to achieve non-contact measurement of the bolt under test, eliminating measurement errors and long-term drift caused by aging, deterioration, or uneven application of acoustic coupling agent in traditional contact measurement, thus improving the long-term stability and reliability of the system. At the same time, its coupling-free characteristic also enables it to adapt to harsh working conditions such as high temperature and vacuum, and provides convenience for automated deployment and inspection, expanding the application scenarios of the system. Attached Figure Description

[0023] Figure 1 This is the functional logic diagram of the integrated axial force measurement and health diagnosis of the present invention;

[0024] Figure 2 This is a flowchart of the online monitoring and multi-mode fault diagnosis process of the system of the present invention;

[0025] Figure 3 This is a diagram illustrating the system architecture for on-site deployment and remote monitoring of the present invention.

[0026] Figure 4 This is a timing diagram of the automatic calibration interaction of the geometric fingerprint template of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0028] The present invention discloses a high-speed real-time ultrasonic axial force measurement system. Its overall architecture mainly includes an electromagnetic ultrasonic transducer module acoustically coupled to the bolt under test, and a core processor module electrically connected to the electromagnetic ultrasonic transducer module and serving as the system control and calculation center. During field deployment, the electromagnetic ultrasonic transducer module is mounted on the end face of the bolt under test using a specially designed positioning fixture to ensure a constant and optimal lift-off distance between the transducer and the bolt surface. This is crucial for ensuring electromagnetic coupling efficiency and long-term stability of the measurement signal. The core processor module is configured to control the electromagnetic ultrasonic transducer module to transmit and receive ultrasonic pulses of two different waveforms in a time-division multiplexing manner, and to perform a series of preset signal processing and calculation procedures on the returned echo signals to ultimately determine the axial force value of the bolt under test and diagnose the bolt's connection status and material health. In a typical... Application scenarios, such as long-term online monitoring of large bridge node bolt groups subjected to alternating cyclic loads and ambient temperature changes, demonstrate the deployment and workflow of this system. In real engineering monitoring scenarios, when a non-uniform temperature gradient forms inside the bolt, due to the influence of sunlight or heat sources from nearby equipment, a non-uniform temperature gradient will form inside the bolt. This causes the axial force measurement method, which relies on external single-point temperature measurement for compensation, to suffer from systematic measurement inaccuracies due to the failure of its physical model assumptions. The core processor module of this invention is configured to execute a measurement procedure based on acoustic feature normalization. Its calculation process mathematically suppresses the influence of uniform temperature changes. Specifically, the core processor module sequentially controls the electromagnetic ultrasonic transducer module to alternately emit a longitudinal wave ultrasonic pulse and a transverse wave ultrasonic pulse into the bolt under test at millisecond intervals. Given that the propagation speed of the longitudinal wave is v... L Simultaneously affected by axial stress σ and temperature T, the propagation speed v of the transverse wave... S Primarily affected by temperature T but insensitive to stress, and both exhibit similar trends in temperature-dependent influence. After receiving the longitudinal and transverse wave echo signals propagating from the bolt, the core processor module calculates an acoustic feature normalization ratio R = v. L / v S Considering that the speed of sound equals twice the effective bolt length L divided by the transit time t, i.e., v L =2L / t L And v S =2L / tS In practice, this normalized ratio is determined as the ratio of the two transit times, R = t. S / t L Thus, the system obtains a characteristic quantity that mainly reflects the change in stress σ, and its calculation results are less affected by uniform temperature changes.

[0029] Another common application-layer obstacle in low-power monitoring systems is that, in order to extract weak echo signals from strong background noise, the signal processing section usually needs to run digital filtering algorithms. This increases the computational burden on the core processor of systems using low-power microcontrollers as the core processor. Therefore, this system adopts a procedure for echo localization based on the geometric characteristics of the bolt itself. The execution of this procedure begins with an initial calibration step. After the bolt under test is first tightened to the target axial force, the core processor module is configured to automatically capture the reflections with stable waveforms generated by the bolt's own fixed geometry, such as threads, chamfers, or steps, in its echo signal. The echo signal is collected and stored in non-volatile memory as a geometric fingerprint template. Considering the energy conversion efficiency fluctuations of the electromagnetic ultrasonic transducer, to ensure stable identification of the echo signal from strong background noise, the core processor module performs a signal averaging algorithm on multiple consecutively acquired echo signals before performing cross-correlation calculations, effectively improving the signal-to-noise ratio. In subsequent routine measurements, the core processor module performs a cross-correlation calculation between the real-time received raw echo signal and the pre-stored geometric fingerprint template. Since the cross-correlation calculation suppresses random noise that does not match the template shape, the peak position of the calculation result can accurately indicate the P-wave transit time t. L and transverse wave transit time t S This procedure utilizes the inherent physical structural characteristics of bolts as the benchmark for signal positioning. Its computational burden is far lower than that of complex filtering algorithms, enabling the entire axial force measurement system to operate for a long time on a low-power, low-cost general-purpose microcontroller.

[0030] After determining the acoustic feature normalization ratio R, the system needs to convert it into the final axial force value. This process is accomplished through a preset mapping relationship, the deterministic procedure of which is as follows: In offline mode, sample bolts of the same batch and specification as the bolt to be tested are selected and installed on a pressure testing machine equipped with a high-precision force sensor; under constant ambient temperature, the testing machine applies an axial load to the sample bolts in incremental steps starting from zero. At each stable load point, the system of this invention measures and records the corresponding acoustic feature normalization ratio R; after completing all measurements from zero load to the maximum expected working load, a set of precisely corresponding [R,F] data pairs can be obtained. The core processor module fits this set of data into a polynomial function or a lookup table and stores it as a preset relationship; in actual field measurement, the core processor module only needs to substitute the real-time calculated R value into the function or interpolate in the lookup table to output the current axial force F value of the bolt; furthermore, the system of this invention... The system not only provides axial force values ​​but is also configured to perform multi-dimensional diagnostics on the service condition of bolts. Under long-term vibration conditions, a failure mode equally critical as the decrease in axial force is abnormal connection conditions such as loose nuts or structural gaps in the connected parts. The system utilizes a byproduct of the cross-correlation process: the residual signal generated by the cross-correlation operation. This residual signal is the remainder after subtracting the matched template signal from the original echo signal. When the bolt connection is in good condition, the energy or information entropy of this residual signal remains at a low baseline level. When abnormalities such as loose nuts occur, new reflection interfaces are introduced into the sound wave propagation path, generating unexpected echoes that cannot be matched by the geometric fingerprint template. These unexpected echoes are retained in the residual signal, causing the real-time calculated energy or information entropy value to be significantly higher than the healthy residual baseline value measured under the initial tightening condition. When the core processor module detects this phenomenon exceeding a preset threshold, such as twice the healthy baseline value, it outputs an independent alarm indicating an abnormal connection condition.

[0031] To provide a supplementary diagnostic path for connection anomalies, the system is also configured to execute a procedure based on acoustic fingerprint comparison. In the initial tightened state, the core processor module controls the transducer to emit a series of coded acoustic packets and records the returned composite echo signals containing reflections from multiple points on the bolt and connection structure, thus establishing a healthy acoustic fingerprint. In subsequent operation, the system periodically emits the same acoustic packets and collects real-time echoes. By comparing the real-time echoes with the healthy acoustic fingerprint using a dynamic time warping algorithm, an alarm indicating an abnormal connection status is output when the calculated waveform similarity is lower than a preset threshold. This method is more sensitive to anomalous patterns that are less likely to generate sharp residual signals, such as gasket creep and minor flange misalignment. Furthermore, for high-strength bolts operating under cyclic loading, the final fatigue fracture begins with the initiation and propagation of microcracks within the material. The core processor module is also configured to perform parallel nonlinear information decoding on the same set of received echo signals to provide early warning of material fatigue damage. Specifically, it performs parallel bandpass filtering on any received echo signal to obtain the amplitude A1 of its fundamental frequency component (e.g., 1 MHz) and the amplitude A2 of its second harmonic component (e.g., 2 MHz). Subsequently, the system calculates a relative nonlinear factor. Provided that the axial force value remains within the preset fluctuation range, the system tracks the β value of the bolt over a long period of time. When its change trend shows a continuous increase and exceeds a damage judgment threshold preset based on material fatigue test data, the system outputs an early warning indicating early material fatigue damage.

[0032] To enhance the reliability and information dimension of nonlinear diagnostics, the system also integrates two in-depth processing procedures. The first is a calibration for the stability of the excitation source. To eliminate the interference of ultrasonic excitation energy changes caused by fluctuations in system power supply voltage or temperature drift of electronic components on long-term β value tracking, the core processor module is configured to additionally extract the amplitude A of an interface reflection wave from the echo signal. interface The interface reflection wave is the first reflected echo generated by the ultrasonic pulse at the interface between the transducer and the bolt. Its amplitude directly reflects the acoustic wave energy injected into the bolt. By utilizing this A... interface The calculated relative nonlinearity factor is normalized, for example, by using... The first aspect addresses the issue of damage location. To suppress interference from excitation energy fluctuations on damage warning at the signal processing level, the core processor module no longer simply extracts the peak amplitude of the second harmonic. Instead, it performs piecewise gradient analysis on the entire echo waveform A2(t) of the second harmonic component along the time axis. Logically, each piece corresponds to a different length segment of the bolt under test, such as the head, middle, and tail. By calculating the average growth slope of the A2(t) waveform within each segment, the segment with the largest average growth slope is identified as the high-risk area where material fatigue damage is most concentrated. Finally, to address the boundary problem of potential accuracy degradation of the main measurement method under thermal gradient conditions, the system also integrates a sound path segmented adaptive correction mechanism. When there is a temperature gradient along the bolt axis, with one end hot and the other cold, the time it takes for the sound wave to travel back and forth multiple times within the bolt will accumulate slightly due to the different temperature regions it passes through. This is specifically reflected in the time interval Δt between two or more consecutive end echoes. echo No longer a constant value; the core processor module calculates the time interval Δt by precisely locating at least two consecutive end echoes. echo Based on the change in this time interval, and according to a pre-defined model based on the thermoacoustic properties of the material, a compensation coefficient K characterizing the severity of the axial temperature gradient is derived. grad Finally, using this compensation coefficient K grad The aforementioned RF mapping relationship can be dynamically modified, for example, adjusted to F = f'(R,K). grad In this invention, the form of a composite piezoelectric transducer is used to compensate for measurement errors caused by the axial temperature gradient, thereby enhancing the applicability of the measurement method under more severe working conditions. Meanwhile, in the description of this invention, the term "composite piezoelectric transducer" is mainly used to refer to the core energy conversion element. However, those skilled in the art should understand that this element may also be referred to as an electromagnetic ultrasonic transducer module in some contexts or according to its function. Therefore, within the context of this invention, the terms "composite piezoelectric transducer" and "electromagnetic ultrasonic transducer module" are used interchangeably, referring to the same component, and both belong to extended embodiments known to those skilled in the art.

[0033] Example 1: In a long-term service scenario of a critical fastening bolt of a nuclear power plant containment vessel, the bolt is simultaneously subjected to cyclic thermal and mechanical loads caused by reactor start-up and shutdown. Its safety status depends not only on maintaining the axial preload but also on two different potential failure modes: material fatigue damage and connection loosening. Conventional monitoring methods can usually only obtain a single axial force-related reading and cannot distinguish the root cause of failure. After the technical solution of this invention is deployed on the bolt, its core processor module has already established the geometric fingerprint template of the bolt according to the above-mentioned procedures during the initial installation stage, and calibrated and stored the baseline value of the residual signal energy and the initial value of the relative nonlinear factor under the healthy fastening state. In the subsequent long-term continuous operation, the system alternately transmits longitudinal and transverse wave pulses at a preset frequency, and continuously calculates the acoustic feature normalization ratio R to output the axial force through cross-correlation calculation with the geometric fingerprint template. At the same time, the echo signal is processed in parallel to monitor the changes in residual signal energy and relative nonlinear factor β.

[0034] After the system had been running for several months, the monitoring data showed two parallel trends: first, the normalized ratio of acoustic features... The calculated axial force value showed a slow decrease, which is consistent with the elastic stress relaxation law of the bolt under this working condition. Secondly, during the same period, the relative nonlinear factor β value calculated by the core processor module after filtering the echo signal showed a continuous and irreversible increasing trend that was unrelated to the gradual change in axial force. At this time, the diagnostic mechanism for connection status in the system provided a key basis for judgment. By performing real-time energy calculation on the residual signal generated by cross-correlation operation, the system confirmed that the energy value of the residual signal always remained near the initially calibrated healthy baseline value without any sudden change. This information indicated that the nut was not loose and there was no gap at the connection interface. Given that the connection status was normal and the axial force changed slowly, the continuous increase in β value was attributed to the microscopic damage caused by the accumulation of cyclic loads inside the bolt material.

[0035] The core processor module then initiated piecewise gradient analysis on the echo waveform A2(t) of the second harmonic component. By calculating the average growth slope of this waveform across different bolt length segments, it was determined that the segment with the largest slope was located at the root of the bolt thread, a known stress concentration area in engineering. This series of technical actions all originated from the same set of acquired echo signals. However, through multi-dimensional information decoding, a signal that might have been interpreted as a single axial force anomaly was transformed into a structured health report containing the current axial force value, connection status assessment, early material damage warning, and location of high-risk damage areas. Based on this report, the operations and maintenance team was able to conduct targeted inspections and replacements of the specific bolt during the next planned outage maintenance window, rather than performing unnecessary emergency outages due to an unexplained axial force anomaly signal. The integrity of the bolt connection structure was maintained, while operational risks that could arise from insufficient decision-making information were avoided.

[0036] Example 2: To quantitatively verify the effectiveness of the technical solution of this invention in suppressing temperature interference and diagnosing material damage, the following experiment was conducted. The purpose of the experiment was to objectively compare the stability of axial force readings under uniform temperature change environment between the acoustic feature normalization measurement method used in this invention and the traditional single longitudinal wave measurement method, and to verify the effectiveness of the relative nonlinear factor in characterizing early fatigue damage of materials. The experiment was conducted on a test platform integrating a high-precision servo loading system and a programmable temperature-controlled environmental chamber. A high-strength bolt of M24 specification and grade 10.9 was selected as the test object, and it was monitored by an electromagnetic ultrasonic transducer module connected to the core processor module of this invention. The experimental data were divided into two groups for comparison. The data processing method of the control group was based solely on the longitudinal wave transit time t. L The change in acoustic characteristics is used to calculate the axial force, simulating the traditional ultrasonic measurement method without effective temperature compensation. However, the data processing method of this invention strictly follows the procedures described in this invention, calculating the acoustic characteristic normalization ratio R = t. S / t L To determine the axial force.

[0037] In the first test, a constant axial preload of 200 kN was applied to the bolt using a servo loading system, the value of which was confirmed by an independent, calibrated force sensor. While maintaining this preload, the temperature of the controlled environmental chamber was cyclically controlled between -10°C and 60°C. This temperature range was designed to simulate the outdoor temperature fluctuations faced by typical industrial and bridge structures during seasonal changes. At each temperature point, sufficient time was allowed for the bolt's internal and external temperatures to reach thermal equilibrium, after which the system collected and recorded a set of measurement data. The test results showed that at -10°C, the control group's reading was 215.8 kN, while... The reading of the sample group of this invention was 201.1 kN. When the temperature rose to 60°C, the reading of the control group dropped to 176.9 kN, with a drift of more than 40 kN within a temperature difference range of 70°C. In contrast, the reading of the sample group of this invention was 198.8 kN at 60°C, and its reading fluctuated within ±1.5 kN around the true axial force of 200 kN throughout the entire temperature range. This result shows that by calculating the ratio of the transit time of the longitudinal and transverse waves, the interference of uniform temperature changes on the measurement results can be suppressed. Its stability stems from the suppression of the trend of the sound velocity of the two sound waves changing with temperature in the ratio calculation.

[0038] In the second experiment, to verify the sensitivity of the relative nonlinear factor β to material fatigue damage, a new bolt of the same specification was selected. At a constant temperature of 25°C, a servo loading system applied an average load of 150 kN, and then superimposed a sinusoidal cyclic load of ±50 kN to accelerate fatigue damage accumulation. At different cycles of cyclic loading, the cyclic loading was paused. Under the static condition of maintaining a 150 kN average load, the system collected echo signals and calculated the relative nonlinear factor β value according to the procedures in the specific implementation method. The experimental results showed that at the initial stage of the experiment, i.e., at 0 cycles, the measured relative nonlinear factor β baseline value was 1.00; when fatigue accumulated to 100,000 cycles, the... The value increased to 1.58; at the end of the test with 200,000 cycles, the β value reached 2.95, while the axial force measured by the sample group method of this invention always fluctuated narrowly between 149.2kN and 150.1kN; the β value showed a clear increasing trend with the increase of fatigue cycles. The underlying mechanism is that the cyclic load caused microcracks to be initiated and extended inside the bolt material. These microscopic nonlinear interfaces will generate second harmonics when disturbed by acoustic waves. The β value, as the ratio of the amplitude of the second harmonic to the square of the amplitude of the fundamental wave, is directly related to the density and size of the nonlinear characteristics inside the material. The test results show that the relative nonlinear factor β can be used as a characteristic parameter to indicate the degree of early fatigue damage accumulation in the material.

[0039] Example 3: This example combines Figures 1 to 4A description of a high-speed real-time ultrasonic axial force measurement system, such as... Figure 1 As shown in the figure, the diagram begins with an electromagnetic ultrasonic transducer module coupled to the bolt under test. This module is configured to alternately transmit and receive longitudinal and transverse waves, and sends the generated raw echo signals to the core processor for further processing. This processing flow is mainly divided into three parallel branches. The first is the core axial force measurement main line, which uses the echo positioning module to accurately determine the longitudinal wave transit time t using a preset geometric fingerprint template and cross-correlation calculation. L With transverse wave transit time t S The system employs several methods: First, it calculates an acoustic characteristic normalization ratio R that suppresses the influence of uniform temperature changes at the principle level. Second, it calculates the axial force value of the bolt based on a preset mapping relationship. Third, it establishes a diagnostic branch for the connection status, which uses the residual signal generated during the cross-correlation operation to analyze unexpected reflections introduced by abnormalities such as loosening, and outputs an alarm for abnormal connection status. Fourth, it establishes a warning branch for material health, which calculates a relatively nonlinear factor that can characterize microscopic damage by decoding the nonlinear acoustic information in the echo signal in parallel, so as to achieve early warning of material fatigue damage.

[0040] like Figure 2 As shown, the process begins with the completion of bolt tightening and installation, followed by a crucial online commissioning and baseline establishment phase to establish the geometric fingerprint and health status baseline necessary for subsequent measurements and diagnostics. After the baseline is established, the system enters a normal monitoring mode for continuous operation. In this mode, the system continuously calculates axial force and monitors various diagnostic indicators. This normal monitoring mode has three main abnormal exit paths, corresponding to three different fault or warning states: First, when the detected signal amplitude or signal-to-noise ratio is lower than the preset standard, the system enters a signal acquisition fault state. Second, when the detected residual signal energy exceeds its health baseline threshold, the system enters a connection status abnormal alarm state. Third, when the detected relative nonlinear factor continues to increase and exceeds its damage judgment threshold, the system enters an early damage warning state. These states can all be manually reset or troubleshooted to return to the normal monitoring process, thus forming a monitoring logic.

[0041] like Figure 3 As shown, in a field environment such as a bridge structural node, multiple high-strength bolts are connected to an electromagnetic ultrasonic transducer module via acoustic coupling. Each ultrasonic transducer is connected to a field monitoring terminal via signal cables. This terminal has embedded measurement and diagnostic firmware. The field monitoring terminal transmits data and status information to a remote monitoring center via a wireless communication network. Inside the center, the data first reaches a data server containing a central database, and then connects to a monitoring workstation via an internal network. This workstation runs data visualization and alarm interface software for professionals to monitor and analyze.

[0042] like Figure 4 As shown in the diagram, the interaction sequence between the internal functional modules is illustrated. The process begins with the operator initiating system initialization. The core processor then enters debug mode. After the operator confirms that the bolt is tightened, the core processor controls the electromagnetic ultrasonic transducer module to emit ultrasonic pulses to the bolt under test. The echo signal returned by the bolt is received by the electromagnetic ultrasonic transducer module and then transmitted to the core processor as echo data. The processor analyzes the data, extracts stable waveform segments, and stores them as geometric fingerprint templates in the storage module. To ensure the validity of the template, the processor automatically executes a self-test loop, clearly marked as being executed 5 times in the diagram. This loop includes steps such as emitting test pulses, receiving echoes, performing cross-correlation verification, and calculating the signal-to-noise ratio. After confirming that the indicators meet the requirements, the processor displays a successful calibration to the operator and finally exits debug mode, transitioning to regular monitoring operation.

[0043] Example 4: In a specific engineering practice, to apply the measurement system of this invention to the critical fastening bolts of the main bearing of a certain type of wind turbine, it is necessary to conduct a systematic offline calibration of its internal calculation model and judgment threshold to ensure that the accuracy of the measurement results and the reliability of the diagnostic function meet the requirements of this specific application scenario. This calibration procedure is performed on an experimental device that integrates a servo hydraulic loading platform local temperature control device and a high-precision data acquisition system. M36 high-strength bolts of the same batch and specification as the bolts to be installed on site are selected as the calibration objects. The first step is to measure the normalized ratio of axial force to acoustic characteristics. To determine the mapping relationship between the samples, the sample bolts were mounted on a loading stage. Under constant temperature conditions of 25°C, an axial tensile force ranging from 0 kN to 500 kN was applied through the loading stage, with the load increasing in 25 kN increments. At each load step, after the fluctuation of the force sensor reading within 10 seconds was less than 0.01% of its full scale, the system of this invention measured and recorded 10 sets of acoustic feature normalized ratios R and took their arithmetic mean. After completing the measurements at all load points, a [R,F] correspondence sequence containing 21 data points was obtained. This sequence was then fitted into a quadratic polynomial F = c²R using the least squares method. 2 +c1R+c0, and store the determined coefficients c2,c1,c0 in the core processor module of the system to be deployed.

[0044] Subsequently, under a constant preload of 200 kN, the axial temperature gradient compensation model was calibrated on the bolt. Using a local temperature control device, a controllable heat source was applied to one end of the bolt, while the other end was kept at room temperature. The temperature difference ΔT at the ends of the bolt was measured using miniature thermocouples placed at both ends. After the temperature difference stabilized at a specific value, the system measured and recorded the time interval Δt between multiple end echoes. echoRepeat this process to obtain Δt under multiple different temperature gradients within the range of 10℃ to 40℃. echo Value, thus establishing a description of Δt echo A data model showing the mapping relationship between ΔT and the compensation coefficient K; this model was used to derive the compensation coefficient K. grad The computational logic is integrated into the core processor module, enabling it to perform calculations based on the real-time measured Δt under field operating conditions. echo The value is used to correct for the influence of the axial temperature gradient; finally, the threshold values ​​related to the diagnostic function are set; for early warning of material fatigue damage, on a new bolt without fatigue loading, 100 independent relative nonlinearity factor β measurements are performed continuously under its rated working load, and the mean value μ of this set of benchmark data is calculated. β With standard deviation σ β The damage assessment threshold for the bolt is set to μ. β +6σ β For diagnosing connection anomalies, a similar statistical method is used: under healthy and secure conditions, the energy values ​​of 100 sets of residual signals are collected, and their mean μ is calculated. E With standard deviation σ E And set the alarm threshold to μ E +5σ E The selection of coefficients is the result of a trade-off between ensuring sensitivity to early loosening and suppressing system electrical noise fluctuations. By executing the above complete calibration and setting procedures, the core processor module of the measurement system is endowed with an experimentally verified calculation model and judgment basis with clear data sources for specific bolt types, thereby enabling it to have traceable measurement accuracy and reliable diagnostic capabilities in subsequent field deployments.

[0045] Example 5: In a field maintenance scenario for a petrochemical pipeline flange, the measurement system of this invention needs to be installed on a replaced connecting bolt. Although the bolt meets specifications, its production batch differs from the sample bolt that underwent offline calibration. Furthermore, the site lacks the necessary large-scale loading equipment. Therefore, a standardized online debugging and baseline establishment procedure needs to be performed simultaneously with the bolt tightening. During the final tightening of the bolt using a torque wrench, the core processor module is placed in a debugging mode. In this mode, the processor acquires ultrasonic echo signals in real time and continuously analyzes the reflected waveform formed by the bolt's geometry. After the solidification operation is completed and the bolt axial force reaches and remains stable, the processor automatically selects the waveform with the highest signal-to-noise ratio and the most stable shape from the echo signal. This waveform usually originates from the transition area between the bolt end thread and the shank, and solidifies it as a geometric fingerprint template for subsequent measurements. To verify the effectiveness of this template, the system immediately performs several rapid self-tests, and confirms through cross-correlation calculations that the signal-to-noise ratio of its positioning peak exceeds a quality threshold set according to the application security level. This threshold is set to 20dB in this maintenance scenario. Only after the verification is passed will the geometric fingerprint template be finally confirmed and used for routine transit time measurements.

[0046] After the geometric fingerprint template is confirmed, and after the on-site engineer completes the tightening operation according to the standard operating procedure and confirms that the connection status is intact, the core processor module automatically executes the diagnostic function baseline establishment procedure; the system performs 50 consecutive measurements, calculates the energy value of the residual signal generated by each cross-correlation operation, and calculates its mean μ based on these 50 data points. E With standard deviation σ E This establishes a localized health residual benchmark and alarm threshold for the bolt in this specific installation state. At the same time, the system controls the transducer to emit a series of sound wave packets and records the returned composite echo signal to establish a health acoustic fingerprint for comparison by the dynamic time warping algorithm. After this series of procedures is completed, the system automatically exits the debugging mode and enters the normal long-term monitoring state, thus completing the functional initialization and baseline establishment on the new bolt.

[0047] Example 6: The diagnostic function of the measurement system of the present invention relies on a set of pre-established baseline models that accurately define the health status of the signal. The establishment procedure of this model is executed after the system is first deployed on a specific bolt and debugged. This procedure aims to establish its own health status parameter boundaries for the signal acquisition link. The process is that the core processor module continuously acquires and analyzes 100 complete sets of raw echo signals after confirming that the signal coupling and system are working normally. For each set of signals, the system first measures the peak amplitude of the first terminal echo and determines an effective amplitude range based on the statistical distribution of these 100 amplitude data. The lower limit of this range is set to 70% of the mean of the data set to characterize the effective boundary of the signal strength. Secondly, the system extracts the initial segment of each signal before any reflected wave arrives, calculates the root mean square value of its background electrical noise, and sets twice the statistical mean of the noise level of these 100 noise level data as the warning threshold for background noise.

[0048] Through the above procedures, the system establishes an online health status baseline model for its signal acquisition link, which includes the lower limit of signal strength, the upper limit of background noise, and corresponding handling logic. In subsequent long-term monitoring, before executing each round of measurement tasks, the system will compare the real-time acquired signal characteristics with this baseline model. If the signal amplitude is lower than the set lower limit or the background noise is higher than the warning threshold, the regular calculation will be interrupted and a diagnostic code indicating an abnormality in the signal acquisition link will be output. This model is the basis for the reliable execution of all subsequent measurement and diagnostic functions.

[0049] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. 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 be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A high-speed real-time ultrasonic axial force measurement system, characterized in that, include: An electromagnetic ultrasonic transducer module is configured to alternately excite longitudinal wave ultrasonic pulses and transverse wave ultrasonic pulses in a bolt under test. A core processor module is connected to an electromagnetic ultrasonic transducer module. The core processor module is configured to: control the electromagnetic ultrasonic transducer module to alternately transmit longitudinal wave ultrasonic pulses and transverse wave ultrasonic pulses to the bolt under test; receive the longitudinal wave echo signal and transverse wave echo signal propagated and returned by the bolt under test via the electromagnetic ultrasonic transducer module; call a geometric fingerprint template established and stored through a calibration process, the geometric fingerprint template being based on the fixed reflection characteristics formed by the geometric structure of the bolt under test, and perform cross-correlation operations between the received longitudinal wave echo signal and transverse wave echo signal and the geometric fingerprint template to determine the transit time of the longitudinal wave and the transit time of the transverse wave; and calculate an acoustic feature normalization ratio based on the determined longitudinal wave transit time and transverse wave transit time. The axial force value of the bolt to be tested is determined based on the mapping relationship between the acoustic characteristic normalization ratio established through a calibration process and the bolt axial force.

2. The ultrasonic axial force high-speed real-time measurement system according to claim 1, characterized in that, The core processor module is configured to calculate the acoustic feature normalization ratio R in the following manner: Among them, v L For the longitudinal wave speed of sound, v S For the speed of sound of a transverse wave, t L Let t be the transit time of the longitudinal wave. S This is the transit time of the transverse wave.

3. The ultrasonic axial force high-speed real-time measurement system according to claim 1, characterized in that, The core processor module is configured to extract a fixed reflected echo waveform with a stable shape generated by the thread and chamfer of the bolt under test after the bolt reaches the target axial force through the initial tightening operation, and store the waveform as a geometric fingerprint template. The cross-correlation operation uses the fixed reflected echo waveform as a reference to lock the position of the same source echo in the subsequently received echo signal.

4. The ultrasonic axial force high-speed real-time measurement system according to claim 1, characterized in that, The core processor module is also configured to: acquire a residual signal generated by cross-correlation operation; calculate at least one parameter value selected from the group consisting of the energy value and information entropy value of the residual signal; compare the parameter value with a health residual benchmark value measured when the bolt under test is in an initial tightened state; and output an alarm indicating an abnormal connection status when the parameter value exceeds the health residual benchmark value.

5. The ultrasonic axial force high-speed real-time measurement system according to claim 1, characterized in that, The core processor module is also configured to: obtain a mismatch value by performing cross-correlation calculation between the quantized echo signal and the geometric fingerprint template; based on a defined functional relationship between the mismatch value and the axial temperature gradient, derive a compensation coefficient characterizing the axial temperature gradient of the bolt under test; and use the compensation coefficient to correct the mapping relationship to compensate for the influence of the axial temperature gradient on the determination of the axial force value.

6. The ultrasonic axial force high-speed real-time measurement system according to claim 1, characterized in that, The core processor module is also configured to filter any received echo signal to obtain the amplitude of its fundamental frequency component and the amplitude of its second harmonic component in parallel. A relative nonlinear factor is calculated based on the amplitude of the fundamental frequency component and the amplitude of the second harmonic component. Under the condition that the axial force value is maintained within the preset fluctuation range, the changing trend of the relative nonlinear factor is monitored. When the changing trend shows a continuous increase and exceeds the damage judgment threshold, an early warning indicating early material fatigue damage is output.

7. The ultrasonic axial force high-speed real-time measurement system according to claim 6, characterized in that, The core processor module is also configured to: additionally monitor the drive current or voltage amplitude of the excitation coil of the electromagnetic ultrasonic transducer module; and use this electrical parameter to normalize the relative nonlinear factor in order to eliminate the interference of excitation energy fluctuations on the early warning.

8. The ultrasonic axial force high-speed real-time measurement system according to claim 6, characterized in that, The core processor module is also configured to: segment the entire echo waveform of the second harmonic component along the time axis, with each segment corresponding to a different length section of the bolt under test; calculate the average growth slope of the second harmonic component amplitude within each section; and identify the section with the largest average growth slope as the high-risk area where material fatigue damage is most concentrated.

9. The ultrasonic axial force high-speed real-time measurement system according to claim 1, characterized in that, The core processor module is a microcontroller with a built-in digital signal processing instruction set. The microcontroller is configured to control the electromagnetic ultrasonic transducer module to complete the time-division alternating transmission of longitudinal wave ultrasonic pulses and transverse wave ultrasonic pulses with millisecond-level timing intervals.

10. The ultrasonic axial force high-speed real-time measurement system according to claim 4, characterized in that, The core processor module is also configured to: when the bolt under test is in its initial tightened state, control the electromagnetic ultrasonic transducer module to emit a series of probe acoustic packets and record the returned composite echo signal containing multiple reflections from the bolt under test and the connection structure to establish a healthy acoustic fingerprint; in subsequent operation, periodically emit the same probe acoustic packets and collect real-time echoes; by comparing the real-time echoes with the healthy acoustic fingerprint using a dynamic time warping algorithm, when the calculated waveform similarity is lower than a preset similarity threshold, an alarm indicating an abnormal connection status is also output.

Images