A Feature Guided Wave Mode Enhancement Method and System Based on the Phase Coherence of Full Matrix Data

Abstract

The present invention discloses a method and system for enhancing eigen-guided wave modes based on the phase coherence of full matrix data; the method includes: arranging a plurality of signal monitoring points in the circumferential direction of the structure to be measured; collecting eigen-guided wave signals and summing them. According to the number of minimum semi-circular phases corresponding to different sampling times, the test coefficients and signal scaling coefficients at different sampling times are obtained. The joint received signal is weighted using the signal scaling coefficients at different sampling times to obtain a mode-enhanced signal. The present invention uses a plurality of signal monitoring points to mutually perform signal excitation and reception, obtains a plurality of independent eigen-guided wave signals, sums them to obtain a joint received signal; and utilizes the differences in the distribution of different eigen-guided wave signals to increase the amplitude of the joint received signal at the moments when the phase distribution is uneven; due to the characteristics that the interference modes and noise phases have an approximately randomly uniform distribution, the present invention can achieve the energy enhancement of the concerned modes with weak dispersion and large group velocity.

Description

Technical Field

[0001] The present invention belongs to the technical field of non-destructive testing and structural health monitoring, and particularly relates to a method and system for enhancing characteristic guided wave modes based on the phase coherence of full matrix data. Background Art

[0002] Special-shaped structures are widely used and are composed of one or more regular or irregular geometric structures, and are usually used in fields such as machinery, aerospace, and architecture. These structures will have a certain degree of structural damage due to various reasons during use or service. Cracks, fatigue damage, deformation, wear, fracture, etc. will be caused by load changes or impacts, and chemical changes caused by environmental factors will lead to structural corrosion, etc. Recent research shows that characteristic guided waves with strong energy aggregation characteristics show good application effects in the defect diagnosis of various special-shaped structures (such as metal welds, bends, etc.). However, the characteristic guided waves in special-shaped structures exhibit multi-modal and dispersive characteristics, which are not conducive to the positioning and evaluation of damage. Considering that the weakly dispersive modes of characteristic guided waves exhibit constructive interference, while other modes and noise exhibit destructive interference, directly summing the amplitudes of the full matrix acquisition data can enhance the weakly dispersive modes and suppress the strongly dispersive modes and noise. However, the effect improvement of the direct amplitude summation method is limited, and under the influence of interfering modes, the signal amplitude of the concerned mode is weak. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for enhancing the modes of characteristic guided waves in special-shaped structures, which uses a commercial finite element software to solve the weakly dispersive and large group velocity frequency ranges of characteristic guided waves, collects the full matrix data of characteristic guided waves in special-shaped structures, and combines the phase coherence index to achieve the energy enhancement of the concerned mode.

[0004] In the first aspect, the present invention provides a method for enhancing characteristic guided wave modes based on the phase coherence of full matrix data, which includes:

[0005] Arranging a plurality of signal monitoring points in the circumferential direction of the structure to be measured;

[0006] Using each signal monitoring point as an excitation point in turn for characteristic guided wave detection. The process of characteristic guided wave detection is: applying an excitation signal at the excitation point, and using the remaining signal monitoring points as receiving points to collect characteristic guided wave signals. Summing all the characteristic guided wave signals to obtain a combined received signal.

[0007] Extracting the phases of each characteristic guided wave signal at different sampling times respectively. Setting a plurality of different semi-circular phase intervals (i.e., 180° phase intervals), for each sampling time, respectively extracting the number of characteristic guided wave signals falling into each semi-circular phase interval and the complementary phase interval, and taking the minimum value as the minimum number of semi-circular phases corresponding to the current sampling time.

[0008] Obtain the test coefficients and signal scaling coefficients at different sampling times according to the number of minimum semi-circular phases corresponding to different sampling times.

[0009] Use the signal scaling coefficients at different sampling times to weight the jointly received signal to obtain a mode-enhanced signal.

[0010] Preferably, the test coefficient P has the following expression:

[0011]

[0012] where is an intermediate solution parameter, and its expression is ; is the number of minimum semi-circular phases.

[0013] Preferably, the signal scaling coefficient is obtained by taking the opposite of the logarithm of the test coefficient P .

[0014] Preferably, the guided wave excitation signal is obtained by screening in a preset frequency range on the condition that the velocity difference between the modes with the fastest and the second-fastest group velocities is the largest.

[0015] Preferably, the preset frequency range is 100 kHz to 150 kHz.

[0016] Preferably, the number of signal monitoring points is greater than 10.

[0017] Preferably, the structure to be measured is a straight rod with a constant cross-section.

[0018] Preferably, the multiple signal monitoring points are arranged at equal intervals in sequence on the edge of the same cross-section of the structure to be measured.

[0019] Preferably, the process of extracting the phase of the characteristic guided wave signal is as follows: perform a Hilbert transform on the characteristic guided wave signal to obtain the imaginary part of the characteristic guided wave signal; calculate the arctangent function of the ratio of the imaginary part to the real part of the characteristic guided wave signal at different sampling times to obtain the phases at different sampling times.

[0020] In a second aspect, the present invention provides a characteristic guided wave mode enhancement system, which is used to execute the foregoing characteristic guided wave mode enhancement method; the characteristic guided wave mode enhancement system includes a signal excitation and acquisition module, a phase extraction module, and a joint weighting module.

[0021] The signal excitation and acquisition module includes a plurality of piezoelectric transducers arranged at each signal monitoring point. The piezoelectric transducers are used to excite and detect characteristic guided waves. The phase extraction module is used to extract the phases of the characteristic guided wave signals at different times. The joint weighting module is used to sum different characteristic guided wave signals, and calculate the signal scaling factors at different times according to the phases of the characteristic guided wave signals, and weight the amplitudes of different times of the joint received signal obtained by summation through the signal scaling factors.

[0022] In a third aspect, the present invention provides a structural damage detection method, and the process is as follows: obtaining a mode enhanced signal of the structure to be measured through the foregoing characteristic guided wave mode enhancement method. Extracting the reflection signal corresponding to the damage from the mode enhanced signal. Obtaining the position of the damage in the structure to be measured according to the arrival time of the reflection signal corresponding to the damage.

[0023] The present invention has the following beneficial effects.

[0024] 1. The present invention uses multiple signal monitoring points to mutually excite and receive signals, obtains a plurality of independent characteristic guided wave signals, sums them to obtain a joint received signal; and utilizes the differences in the distribution of different characteristic guided wave signals to increase the amplitude of the joint received signal at the moments when the phase distribution is uneven; due to the characteristics that the interference modes and noise phases have an approximately random and uniform distribution, the present invention realizes the energy enhancement of the concerned modes with weak dispersion and large group velocity.

[0025] 2. The amplified concerned mode wave in the present invention contains information such as propagation time and waveform, so that it can more conveniently distinguish the arrival time of the reflection signal at the structural damage position, and realize more accurate detection of the structural damage position.

[0026] 3. The process of enhancing the characteristic guided wave mode in the present invention does not depend on an accurate characteristic guided wave dispersion curve, and the operation is more convenient and efficient. Description of the Drawings

[0027] Figure 1 It is a flowchart of Embodiment 1 of the present invention;

[0028] Figure 2 It is the guided wave group velocity curve in the T-shaped rod to be measured in Embodiment 1 of the present invention;

[0029] Figure 3 It is the signal monitoring point distribution diagram of the cross section of the T-shaped rod to be measured in Embodiment 1 of the present invention;

[0030] Figure 4 It is the waveform diagram of part of the full matrix acquisition signals in Embodiment 1 of the present invention;

[0031] Figure 5Comparison diagram of the results of Embodiment 1 (weighting after summing simulation signals) and Comparative Example 1 (direct summing of simulation signals) of the present invention;

[0032] Figure 6 Waveform diagram of partial full matrix acquisition signals in Embodiment 2 of the present invention;

[0033] Figure 7 Comparison diagram of the results of Embodiment 2 (weighting after summing experimental signals) and Comparative Example 2 (direct summing of experimental signals) of the present invention. Detailed implementation manners

[0034] The present invention will be further described below with reference to the accompanying drawings.

[0035] To better understand the technical solution of the present invention, three embodiments are respectively used for description. In Embodiment 1, a T-shaped rod structure is established by means of finite element software, and monitoring points are virtually set to complete signal simulation. The amplitudes of the simulation signals are summed, the inspection coefficient is calculated based on the signal phase, and finally the weighted modal signal is calculated using the summed amplitude and the inspection coefficient. In Embodiment 2, a T-shaped rod in kind is directly used, the received signal is obtained through experiments, and the modal enhancement result is calculated by the processing method in Embodiment 1 to verify the rationality of the present invention. In Embodiment 3, a groove or a small hole is cut in the T-shaped rod in kind to represent a damage model, the received signal is obtained through the experimental method in Embodiment 2, and the modal enhancement result is calculated for damage location.

[0036] Embodiment 1

[0037] A method for enhancing characteristic guided wave modes based on the phase coherence of full matrix data, which is carried out by simulation, and the detection object is a T-shaped rod; the T-shaped rod has a T-shaped cross-section as shown in Figure 3 In other embodiments, the structure to be measured may also be other straight rods (specifically, rods with a constant cross-sectional size and shape and a straight axis), or structures other than other straight rods.

[0038] The method for enhancing characteristic guided wave modes includes the following steps:

[0039] S1: Calculate the group velocity characteristics of each mode of the guided wave of the T-shaped rod to determine the center frequency of the guided wave excitation signal.

[0040] S11: The design dimensions of the T-shaped rod structure are as shown in Figure 3 The length of the T-shaped rod is 500 mm, the transverse width is 55 mm, the longitudinal height is 36 mm, and the wall thickness is 5 mm. The mass density of the T-shaped rod is 2700 kg / m 3 , the Young's modulus is 70 GPa, and the Poisson's ratio is 0.3.

[0041] S12: Calculate the group velocity characteristics of each mode of the guided wave of the T-shaped rod by the semi-analytical finite element method as shown inFigure 2 as shown in Figure 2 It contains 35 velocity curves of different modes; select the center frequency of the appropriate guided wave excitation signal according to the number of modes and the velocity difference. In this embodiment, considering the frequency response characteristics of the sensor, select the frequency with the largest velocity difference between the modes with the fastest and the second-fastest group velocities between 100 kHz and 150 kHz as the center frequency of the guided wave excitation signal.

[0042] To select the characteristic guided wave detection frequency, define the frequency selection factor as follows

[0043] Equation (1)

[0044] where represents the velocity difference between the modes with the fastest and the second-fastest group velocities at frequency f; represents the maximum group velocity at frequency f.

[0045] From Figure 2 it can be seen that when the frequency is 125 kHz, the frequency selection factor has a relatively large value. Therefore, select 125 kHz as the center frequency of the guided wave excitation signal. When the center frequency of the guided wave excitation signal is 125 kHz, the maximum group velocity is about 4500 m / s. Therefore, the time taken for the guided wave to travel from transmission to reception is about 0.22 ms.

[0046] This step S1 can obtain the frequency range with weak dispersion and large group velocity as the frequency of the excitation signal.

[0047] S2: Use Abaqus CAE finite element software to establish a T-shaped rod structure model and perform a display dynamics analysis to generate simulation signals.

[0048] S21: Use a cube with a side length of 0.5 mm as the mesh division unit.

[0049] S22: Uniformly and symmetrically create 24 groups of nodes ( ) on the cross-section of the T-shaped rod, that is, 24 signal monitoring points, and their monitoring point distribution is as Figure 3 shown. The signal monitoring points are arranged at the ends of the T-shaped rod.

[0050] S23: Use each node as the excitation point in turn and perform characteristic guided wave propagation simulations respectively.

[0051] The process of the characteristic guided wave propagation simulation is: use all nodes except the excitation point as the receiving points; apply a sine excitation signal with 12 cycles and a center frequency of 125 kHz to the excitation point, and the effective simulation time is 0.5 ms. After submitting the completion of the simulation, obtain the initial signal waveform of the simulation and the received signals at each receiving point , where i is the node number of the excitation point and j is the node number of the receiving point; therefore, 23 received signals are obtained for each characteristic guided wave propagation simulation.

[0052] After completing the characteristic guided wave propagation simulations with all nodes as excitation points, the complete simulation received signals are obtained, denoted as ( ), ( t∈[0, 5×10 -4 s] ); for the convenience of analysis, directly use to represent the 552 full matrix received signals ( ).

[0053] S3: Extract the phase of the received signal.

[0054] S31: Discretize the simulation received signal. Divide the simulation duration of the simulation received signal into 10,000 unit steps , ; denote different moments ( ), and the amplitude and phase of each simulation received signal at moment are denoted as and ( ).

[0055] S32: Analyze all the simulation received signals to obtain the analytic signal as follows:

[0056] Equation (2)

[0057] In the formula, is the imaginary unit; is the amplitude of the discrete simulation received signal; is the amplitude obtained by Hilbert transform of the imaginary part.

[0058] The process of calculating the imaginary part is as follows: Perform Hilbert transform on the simulation received signal before discretization to obtain the transformed signal of the continuous signal; then discretize the transformed signal to obtain the imaginary part of the discrete real-time analytic signal; the Hilbert transform formula is as follows:

[0059] Equation (3)

[0060] S32: Extract the phase of the real-time signal. Based on the analytic signal , the phases of the simulated received signal at different times can be obtained:

[0061] Equation (4)

[0062] S4: Calculate the amplitude summation and test coefficient of the simulated signal.

[0063] S41: Sum the amplitudes of the discrete simulated received signals to obtain the combined received signal , and the expression is as follows:

[0064] Equation (5)

[0065] Where , .

[0066] S42: Calculate the test coefficient based on the signal phase.

[0067] Establish the null hypothesis and the alternative hypothesis as follows:

[0068] Null hypothesis : The phases of the simulation signals are uniformly distributed on the circumference;

[0069] Alternative hypothesis , The phases of the simulation signals are not uniformly distributed on the circumference;

[0070] Take the probability of the null hypothesis H0 being true as the test coefficient P .

[0071] Establish the general formula for the test coefficient P as follows:

[0072] Equation (6)

[0073] Where represents the number of received signals. Since the phases of the received signals , The phases of the signals are distributed on a circle. Intercept several semi-circles from this circular system; is the number of minimum semi-circle phases of the full matrix received signal at time .

[0074] Since the number of signals at each moment , in order to improve the calculation efficiency, the calculation check coefficient P is simplified by the following formula :

[0075] Equation (7)

[0076] where is the intermediate solution parameter, and its expression is ;

[0077] The minimum number of semi - circle phases is solved as follows:

[0078] In the circles composed of the real - time phases of all simulation signals corresponding to different times, 180 semi - circles are intercepted from each circle by the system, and the obtained semi - circle intervals are respectively , , , ; Among the semi - circles and their complementary semi - circles corresponding to the circles at different times, find the number of phases of the semi - circle with the least number of phases falling in each circle, which is , and the specific calculation process is as follows:

[0079] For any semi - circle, a phase identifier is established for the phase of each simulation received signal as follows:

[0080] , ( , ) Equation (8)

[0081] where is the phase of the simulation received signal; is the serial number of the semi - circle.

[0082] Calculate the number of phases in each semi - circle as follows:

[0083] Equation (9)

[0084] By taking the minimum value of the number of phases in all semi - circles and the number of phases in their complementary semi - circles, the minimum number of semi - circle phases is obtained as follows:

[0085] Equation (10)

[0086] For the combined received signal at any time , the corresponding check coefficient P The smaller it is, the smaller the probability that the original hypothesis holds, The more the phases of the received simulation signals do not conform to the circular uniform distribution, the greater the probability of phase coherence at the current moment, that is, the greater the probability of the appearance of the characteristic guided wave concerned mode; therefore, the test coefficient P is smaller, which indicates that the signal values at any moment should be amplified more. Therefore, in the subsequent step S5 of this embodiment, the opposite of the logarithm of the test coefficient P is taken, and the obtained value is multiplied by the combined received signal at any moment to enhance the amplitude of the effective component.

[0087] S5: Calculate the signal scaling coefficients at different moments according to the test coefficient P at different moments as follows:

[0088] Equation (11)

[0089] Use the signal scaling coefficients at different moments to weight the combined received signal to obtain the modal signal after enhancing the characteristic guided wave mode as follows:

[0090] Equation (12)

[0091] Figure 4 is the waveform of some full matrix acquisition signals obtained by simulation in this embodiment. It can be seen from the figure that the end face emission wave of each modal signal appears at about 0.22 ms, that is, the end face reflection wave. The length of the T-shaped rod is 500 mm. From this, the propagation speed of the fastest mode of the characteristic guided wave in the T-shaped rod can be calculated to be about 4500 m / s, and the excitation signal frequency is 125 kHz, which basically conforms to Figure 2 the group velocity model in

[0092] Comparative Example 1

[0093] A method for enhancing the characteristic guided wave mode; the difference between this comparative example and Embodiment 1 is only that steps S42 and S5 are not executed; the combined received signal directly obtained by summing the amplitudes of the simulation received signals is used as the enhanced signal.

[0094] Figure 5 is the comparison diagram of the modal signals of direct summation and summation weighting after normalization processing in this embodiment. It can be clearly seen that the end face reflection wave appears at about 0.22 ms by direct summation. Compared with the directly summed signal, the amplitude of the end face reflection wave of the summed and weighted signal is significantly enhanced, while the amplitudes of the direct wave, end face reflection wave, and interference mode of the directly summed signal have small differences, and the amplitude of the concerned mode is low.​

[0095] Figure 5 It is a comparison diagram of the signals obtained from Example 1 (weighting after summing simulation signals) and Comparative Example 1 (directly summing simulation signals). In the signal obtained by directly summing in Comparative Example 1, it can be clearly seen that an end-face reflection wave appears around 0.22 ms. Compared with Comparative Example 1 of the direct summation scheme, the amplitude of the end-face reflection wave of the signal obtained in Example 1 of the summation weighting scheme is significantly enhanced. For the directly summed signal, the amplitudes of the direct wave, end-face reflection wave, and interference modes have small differences, and the amplitude of the concerned mode is low.

[0096] Example 2

[0097] A method for enhancing characteristic guided wave modes based on the phase coherence of full matrix data, which is carried out through experiments to verify the feasibility of the present invention. The specific steps of this example are as follows:

[0098] S1: Arrange 24 monitoring points on the cross-section of the T-shaped rod according to the description of Example 1. The monitoring points use piezoelectric ceramic patches. An excitation is applied to each monitoring point respectively, and the remaining 23 monitoring points receive. The excitation source uses a windowed sine signal, and the oscilloscope collects the full matrix data.

[0099] S2: Analyze the amplitude and phase of the signal. Use the method in step S4 of Example 1 to sum the amplitudes of the experimental signals at different times, and calculate the test coefficients at different times based on the phase; use the method in step S5 of Example 1 to weight different times of the signal obtained by summation using the test coefficients at different times to obtain the signal waveform.

[0100] Figure 6 It is the waveform of part of the full matrix acquisition signals obtained through the experiment in Example 2. The propagation time of its end-face reflection wave is basically the same as the software simulation result, both around 0.22 ms, verifying the Figure 3 group velocity model in.

[0101] Comparative Example 2

[0102] A method for enhancing characteristic guided wave modes; the difference between this comparative example and Example 1 is only that: directly use the signal obtained by summing the amplitudes of the experimental signals as the enhanced signal.

[0103] Figure 7Figure for comparing the signals obtained in Example 2 (weighting after summing experimental signals) and Comparative Example 2 (direct summing of experimental signals). In the signal obtained by direct summing in Comparative Example 2, a relatively strong end-face reflection wave can be seen, but its amplitude is significantly lower than that of the direct wave experimental signal. This is because there is strong crosstalk near the initial moment. Compared with Comparative Example 2 of the direct summing scheme, Example 2 of the summing and weighting scheme can, to a certain extent, suppress the direct wave and interference modes of the signal and enhance the amplitude of the concerned mode.

[0104] Example 3

[0105] A structural damage detection method for detecting the damage position on the axial direction of a rod. The detection object is a damaged T-shaped rod with a groove or crack of a certain size at a certain position.

[0106] The process of this structural damage detection method is as follows:

[0107] According to the method of Example 2, 24 monitoring points are arranged and recorded on the cross-section of the damaged T-shaped rod to be measured. Excitations are added in sequence, and the other monitoring points receive the damage experimental signals. The damage experimental signals are processed by extracting the phase, summing the amplitudes, calculating the test coefficient based on the phase, and weighting to obtain a waveform containing the damage reflection signal.

[0108] In the obtained reflection signal waveform, the damage reflection wave appears at a position between the direct wave and the end-face reflection wave. Due to the damage of the T-shaped rod, when the guided wave propagates to the damage position, it is reflected back, thus generating a damage reflection wave. From Example 2, it can be obtained that under the condition of no damage, the propagation time of the guided wave in the T-shaped rod , the rod length is , the axial position of its damage is denoted as , the appearance time is denoted as , thus obtaining the following relationship:

[0109] Equation (13)

[0110] From this, the axial damage position of the T-shaped rod is obtained as follows:

[0111] Equation (14)

[0112] where the time t and the time t' are both obtained by reading from the reflection signal waveform. The rod length is obtained by measurement.

[0113] In this example, the damaged T-shaped rod is obtained by cutting a groove or crack at a preset position on the T-shaped rod in advance, thereby judging the accuracy of the structural damage detection method.

Claims

1. A method for enhancing characteristic waveguide modes based on full matrix data phase coherence, characterized in that: The method includes: Arrange multiple signal monitoring points in the circumferential direction of the structure under test; Each signal monitoring point is used as an excitation point in turn to perform characteristic waveguide detection; the process of characteristic waveguide detection is: applying an excitation signal at the excitation point, using the remaining signal monitoring points as receiving points, and collecting characteristic waveguide signals; summing all characteristic waveguide signals to obtain a joint receiving signal; Extract the phase of each characteristic waveguide signal at different sampling moments respectively; set a plurality of different semicircular phase intervals, and for each sampling moment, extract the number of characteristic waveguide signals falling into each semicircular phase interval and complementary phase interval respectively, and take the minimum value as the minimum semicircular phase number corresponding to the current sampling moment; According to the minimum number of semicircular phases corresponding to different sampling moments, the test coefficient and signal scaling coefficient at different sampling moments are obtained; The joint received signal is weighted using signal scaling coefficients at different sampling times to obtain a modal enhancement signal.

2. The characteristic waveguide mode enhancement method according to claim 1, characterized in that: The test coefficient P The expression is: ； in, is the intermediate solution parameter, and its expression is ; is the minimum number of semicircular phases.

3. The characteristic waveguide mode enhancement method according to claim 2, characterized in that: The signal scaling factor is obtained by the test coefficient P Take the opposite of the logarithm to get .

4. The characteristic waveguide mode enhancement method according to claim 1, characterized in that: The waveguide excitation signal is obtained by screening in a preset frequency interval according to the maximum speed difference between the mode with the fastest group velocity and the mode with the second fastest group velocity as a condition.

5. The characteristic waveguide mode enhancement method according to claim 4, characterized in that: The preset frequency range is 100kHz to 150kHz.

6. The characteristic waveguide mode enhancement method according to claim 1, characterized in that: The number of the signal monitoring points is greater than 10.

7. The characteristic waveguide mode enhancement method according to claim 1, characterized in that: The multiple signal monitoring points are arranged in sequence and at equal intervals on the edge of the same cross section of the measured structure.

8. The characteristic waveguide mode enhancement method according to claim 1, characterized in that: The process of extracting the phase of the characteristic waveguide signal is as follows: performing Hilbert transform on the characteristic waveguide signal to obtain the imaginary part of the characteristic waveguide signal; calculating the inverse tangent function of the ratio of the imaginary part to the real part of the characteristic waveguide signal at different sampling times to obtain the phase at different sampling times.

9. A characteristic waveguide mode enhancement system, characterized in that: Used to execute the characteristic waveguide mode enhancement method as claimed in claim 1; the characteristic waveguide mode enhancement system includes a signal excitation acquisition module, a phase extraction module, and a joint weighting module; The signal excitation acquisition module includes a plurality of piezoelectric transducers arranged at each signal monitoring point; the piezoelectric transducer is used to excite and detect characteristic waveguides; the phase extraction module is used to extract the phases of characteristic waveguide signals at different times; the joint weighting module is used to sum different characteristic waveguide signals, and calculate the signal scaling coefficients at different times according to the phases of the characteristic waveguide signals, and weight the amplitudes of the summed joint received signal at different times by the signal scaling coefficient.

10. A method for detecting structural damage, characterized in that: The method is as follows: obtaining a modal enhancement signal of the structure under test by the characteristic guided wave modal enhancement method as described in claim 1; extracting a reflection signal corresponding to the damage from the modal enhancement signal; and obtaining the location of the damage in the structure under test according to the arrival time of the reflection signal corresponding to the damage.

Images