Method for detecting corrosion of a sheet metal part

Abstract

The present application relates to the field of mechanical structure health monitoring, in particular to a kind of plate corrosion detection method, comprising the following steps: the parameter of plate to be measured is substituted into frequency equation to calculate the phase velocity and group velocity dispersion curve of multi-mode guided wave;According to the slope of the group velocity dispersion curve of each mode guided wave, its corrosion sensitivity can be determined, select two working modes and determine the working frequency of each;Broadband excitation is applied to the plate to generate multi-mode guided wave, and the guided wave is detected at the edge of the plate to be detected area, the time of flight of the selected two working modes at the respective working frequency is obtained;The ratio of the time of flight of the two working modes at the respective working frequency is calculated;The grid is divided in the area to be detected, and the defect distribution probability of the grid is obtained according to the position relationship between the corresponding grid and each propagation path and the corrosion index of each propagation path.The present application solves the safety accidents and losses caused by corrosion, realizes the detection of plate corrosion, and improves the safety of equipment.

Description

TECHNICAL FIELD

[0001] The present application relates to a kind of plate corrosion detection method, belong to mechanical health monitoring field. BACKGROUND

[0002] Corrosion is one of the important damage forms of equipment and components, which, together with fracture and wear, constitutes the three major damage forms of materials in use, often leading to early failure or accidents, causing huge losses. Corrosion is always "in progress", in thermodynamics, except for a few noble metals (Au, Pt, etc.), general materials corrosion is a spontaneous process. Corrosion cracking causes "running, leaking, dripping and leaking", not only wastes resources, energy and pollutes the environment, but also causes fire and explosion accidents. Under the action of corrosive medium, the material will lead to wall thickness thinning, cracking and material structure change, which will reduce the mechanical properties of the material, cause the equipment to break due to insufficient bearing capacity, and thus cause safety accidents. SUMMARY

[0003] The purpose of the present application is to provide a kind of plate corrosion detection method, to solve the problem of plate corrosion detection.

[0004] To achieve the above purpose, the scheme of the present application includes:

[0005] The present application provides a kind of plate corrosion detection method, including the following steps:

[0006] 1) The density, Young's modulus, Poisson's ratio and thickness parameters of the plate to be measured are substituted into the frequency equation to calculate the phase velocity and group velocity dispersion curve of multi-modal guided wave;

[0007] 2) The slope of the group velocity dispersion curve of each mode guided wave can determine its sensitivity to corrosion damage, select two working modes and determine their respective working frequencies;

[0008] 3) A broadband excitation is applied to the plate to generate multi-modal guided waves, and the guided waves are detected at the edge of the plate to be detected, and the time of flight of the selected two working modes at their respective working frequencies is obtained;

[0009] 4) Calculate the ratio of the time of flight of the two working modes at their respective working frequencies, which is the corrosion index of the corresponding propagation path in the area to be detected;

[0010] 5) Divide the area to be detected into a grid, and according to the position relationship between the corresponding grid and each propagation path and the corrosion index of each propagation path, obtain the defect distribution probability of the grid.

[0011] The beneficial effect is that the application provides a plate corrosion detection method. By applying a wideband excitation to the plate, multi-modal guided waves can be generated. Since corrosion reduces the thickness of the plate, the thickness of the plate changes the transmission speed of the guided waves in the plate, and the change in thickness has different effects on the transmission speed of different modal guided waves, so the application can estimate the corrosion position by the speed change ratio of different modal guided waves. By detecting the guided waves at the edge of the plate detection area, the time of flight of the guided waves can be obtained. Since the flight distance of the guided waves is fixed, by calculating the ratio of the time of flight of different modal guided waves, the speed change ratio of different modal guided waves can be determined, and the corrosion index of the corresponding detection area on the propagation path can be obtained. Further, the detection area is divided into a grid, and the defect probability of each grid is calculated by the position relationship of each grid relative to the propagation path and the corrosion index of each grid, so that the corrosion position can be predicted.

[0012] Further, in step 1), the phase velocity dispersion curve of the guided wave is calculated by the Rayleigh-Lamb frequency equation, and the symmetric mode satisfies

[0013]

[0014] The antisymmetric mode satisfies:

[0015]

[0016] The calculation formula of the parameter p is The calculation formula of the parameter q is ω is the angular frequency / rad, c L is the longitudinal wave speed, k is the wave number, c T is the transverse wave speed, and d is the fixed plate thickness.

[0017] The beneficial effect is that the symmetric mode and the antisymmetric mode are calculated by the above formula.

[0018] Further, in step 3), the guided wave time of flight is the difference between the guided wave arrival time and the emission time; the guided wave arrival time is the peak value of the envelope curve, wherein the envelope curve is obtained by Hilbert transform on the narrowband response of the selected guided wave mode; and the guided wave mode includes the symmetric mode and the antisymmetric mode.

[0019] The beneficial effect is that by Hilbert transform on the narrowband response of the guided wave mode, the envelope curve can be obtained, the peak value of the envelope curve is taken as the corresponding arrival time, the arrival time of the guided wave can be obtained, and the calculation of the guided wave time of flight is facilitated, and the calculation of the group velocity is further completed.

[0020] Further, the guided wave modes are selected according to the sensitivity to corrosion, the sensitivity of the guided wave modes to corrosion is determined by the amount of change in the guided wave velocity caused by the change in the thickness of the plate, and two guided wave modes with high sensitivity to corrosion and opposite group velocity change trends under the condition of thinning of the structure thickness are selected as working modes.

[0021] The beneficial effect is that when selecting the guided wave modes, two guided wave modes with high sensitivity to the change in the thickness of the plate and opposite group velocity change trends under the condition of thinning of the structure thickness are selected, which means that the influence of corrosion on the guided wave velocities of the two selected modes is more obvious, the ratio deviation of their flight times is larger, and the detection result is more accurate and has higher precision. Therefore, when selecting the guided wave modes, multiple data tests of the symmetric mode and the anti-symmetric mode are required to select more suitable guided wave modes.

[0022] Further, when selecting the guided wave modes, the change in the guided wave velocity is determined by monitoring the change in the flight time of the wave packet from the excitation position to the receiving position.

[0023] The beneficial effect is that when selecting the modes, it is difficult to calculate the change in the velocity of the guided wave, and therefore the change in the flight time of the guided wave is calculated to reflect the change in the velocity of the guided wave.

[0024] Further, the formula for calculating the change in the flight time is:

[0025]

[0026] The beneficial effect is that the change in the flight time is obtained, and the change in the velocity can be more accurately calculated according to the above formula, so that the selected guided wave modes can be selected.

[0027] Further, the response of the symmetric mode is:

[0028] u s (t)=∫E(ω)H S (ω)e iωt dω=∫E(ω)[H u (ω)+H b (ω)]e iωt dω

[0029] wherein E(ω) is the Fourier transform of the response e(t) of the symmetric mode, H u (ω) and H b (ω) are system transfer functions estimated according to the receiving signals of the upper and lower transducers, H S (ω) is the symmetric mode.

[0030] The response of the anti-symmetric mode is:

[0031] u a (t)=∫E(ω)H A (ω)e iωt dω=∫E(ω)[H u (ω)-H b (ω)]e iωt dω

[0032] Where E(ω) is the Fourier transform of the antisymmetric mode response e(t), and H u (ω) and H b (ω) are the system transfer functions estimated from the received signals of the upper and lower transducers, respectively. A (ω) represents the antisymmetric mode.

[0033] The beneficial effect is that the above formula can be used to calculate the symmetric and antisymmetric modal responses.

[0034] Furthermore, the corrosion index CI n The calculation method is as follows:

[0035]

[0036] Among them, t tof1 The time of flight of the guided wave in the first operating mode is t. tof2 Let L be the flight time of the guided wave in the second operating mode, L be the length from the excitation position to the receiving position, and l be the length of the guided wave through the eroded region. The group velocity of the first working mode at the corroded site. The group velocity of the first working mode in the uncorroded area. The group velocity of the second working mode at the corroded site. This represents the group velocity of the second working mode in the uncorroded area.

[0037] The beneficial effect is that by obtaining the flight time of different modes, the group velocity through the corroded region, and the group velocity through the uncorroded region, the corrosion index CI can be completed. n The calculation.

[0038] Furthermore, the method for calculating the relative sensing path is as follows:

[0039]

[0040] Among them, (x n1 ,y n1 (x) represents the excitation position coordinates. n2 ,y n2 () represents the coordinates of the detection location.

[0041] The beneficial effect is that the position relationship between the corresponding detection path and the grid is reflected by detecting the position relationship between the start and end points of the path and the current grid, and the position relationship is accurate and easy to calculate.

[0042] Further, the probability of grid defect occurrence is:

[0043]

[0044] Wherein, β is a scale parameter for controlling the size of the elliptical distribution area.

[0045] The beneficial effect is that the probability of corrosion occurrence of each grid is calculated according to the position relationship between different grids and each detection path, and the corrosion index of each detection path. The detection path is close to the corresponding grid, and the corrosion index is high, so the probability of corrosion occurrence in the corresponding grid is high. The method of the application realizes the corrosion detection of the plate, has high accuracy and is easy to implement. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 is a schematic diagram of sensor arrangement;

[0047] Figure 2 is a group velocity dispersion curve of an aluminum plate;

[0048] Figure 3 is a curve diagram of the relationship between the velocity change amount and the frequency after the thickness of the corroded plate is reduced from 3.9mm to 3.85mm;

[0049] Figure 4 is a curve diagram of the change amount of the time of flight of each mode of the guided wave through a certain sensing path of the corrosion;

[0050] Figure 5(a) is a reconstruction schematic diagram under S0 mode 400kHz and A1 mode 460kHz;

[0051] Figure 5(b) is a reconstruction schematic diagram under S0 mode 400kHz and A1 mode 480kHz;

[0052] Figure 5(c) is a reconstruction schematic diagram under S0 mode 400kHz and A1 mode 500kHz;

[0053] Figure 5(d) is a reconstruction schematic diagram under S0 mode 400kHz and A1 mode 520kHz;

[0054] Figure 5(e) is a reconstruction schematic diagram under S0 mode 360kHz and A1 mode 500kHz;

[0055] Figure 5(f) is a reconstruction schematic diagram under S0 mode 380kHz and A1 mode 500kHz;

[0056] Fig. 5(g) is a schematic diagram of reconstruction under S0 mode 400 kHz and A1 mode 500 kHz;

[0057] Fig. 5(h) is a schematic diagram of reconstruction under S0 mode 420 kHz and A1 mode 500 kHz. DETAILED DESCRIPTION

[0058] The application will be further described in detail below with reference to the accompanying drawings:

[0059] Lamb wave is one of ultrasonic guided waves, which has the characteristics of high detection efficiency, weak attenuation, and sensitivity to structural damage, and is considered as the most potential method for detecting plate and shell structure damage. In the implementation process of the application, a sensor array is used to obtain the propagation signal of Lamb wave in the structure, and the detection and evaluation of corrosion damage are realized by extracting the characteristic parameters sensitive to faults.

[0060] In order to achieve accurate detection of corrosion damage, the embodiment provides a corrosion detection method for a plate, which comprises the following steps:

[0061] 1) Sensor installation.

[0062] The detection area is planned, the coordinate system is established, the grid is divided as the detection pixel point coordinate, the sensor array is arranged on the component to be detected, and the sensor position coordinate is recorded.

[0063] 2) Mode control.

[0064] According to the displacement direction and motion of the mass point, the guided wave can be divided into symmetric mode (S mode) and anti-symmetric mode (A mode). In particular, the symmetric mode is mainly in-plane displacement of the mass point, which leads to expansion and contraction of the plate along the thickness direction. On the contrary, the anti-symmetric mode is mainly off-plane displacement of the mass point, which leads to bending of the plate without changing the thickness. By placing two identical ultrasonic receiving probes on the upper and lower sides of the same position of the aluminum plate, the signals recorded by the upper and lower probes are arithmetically operated to realize mode control. Adding these signals can enhance the symmetric mode and weaken the anti-symmetric mode. On the contrary, subtracting these signals can enhance the anti-symmetric mode and suppress the symmetric mode.

[0065] 3) Selecting operation point.

[0066] The accuracy of corrosion detection depends on the change of velocity of guided wave propagating through the thinned area. Therefore, the selected mode must have significant dispersion. On the other hand, the measurement of guided wave arrival time must be accurate enough. However, the energy of wave packet spreads in time domain due to dispersion, which leads to change of wave shape and inaccurate time of flight estimation. Therefore, narrow-band windowed toneburst is usually selected as input signal instead of short sharp pulse. For narrow-band detection, the mode and frequency of guided wave determine the accuracy of corrosion detection.

[0067] Wave propagation in a plate with large product of frequency and thickness (including many modes of wide frequency band) can carry rich structural characteristic information for defect detection. A number of narrowband waveguide responses can be extracted from it, whose frequency bands are contained in the wideband excitation frequency band range, which can be an effective method to optimize waveguide detection. The present invention determines the appropriate operating point from the corrosion sensitivity.

[0068] The waveguide mode sensitivity to corrosion is defined as the amount of velocity change (increase or decrease) caused by a given change in plate thickness. In particular, the plate thickness changes from d to d', the amount of velocity change can be calculated as follows:

[0069] velocity variation = |c g (fd)-c g (fd')| (1)

[0070] Therefore, the sensitivity is directly related to the degree of dispersion of the selected operating point. The steeper the slope of the group velocity dispersion curve, the higher the sensitivity of the mode to the defect.

[0071] In practical applications, the amount of velocity change is determined by monitoring the change in the time of flight required for a wave packet to propagate from the excitation to the reception. If the sensing path passes through a length of the corrosion area l, the amount of change in the time of flight is calculated as follows:

[0072]

[0073] where L is the length from the excitation position to the reception position, l is the length of the waveguide passing through the corrosion area, c g (fd') is the group velocity of the corrosion site, c g (fd) is the group velocity of the uncorroded site.

[0074] The formula for calculating the time of flight ratio is:

[0075]

[0076] where t tof1 is the time of flight of the waveguide of the first working mode, t tof2 is the time of flight of the waveguide of the second working mode, L is the length from the excitation position to the reception position, l is the length of the waveguide passing through the corrosion area, is the group velocity of the first working mode at the corrosion site, is the group velocity of the first working mode at the uncorroded site, is the group velocity of the second working mode at the corrosion site, is the group velocity of the second working mode at the uncorroded site.

[0077] By considering the corrosion sensitivity, the selected guided wave mode and frequency can be determined.

[0078] 4) Calculation results.

[0079] After the optional operating point (guided wave mode and frequency) is determined, the corrosion can be detected using a narrowband windowed toneburst excitation. The response of the symmetric mode is calculated according to equation (3):

[0080] u s (t) = ∫ E(ω) H S (ω) e iωt dω = ∫ E(ω) [H u (ω) + H b (ω)] e iωt dω (3)

[0081] where E(ω) is the Fourier transform of the response e(t) of the symmetric mode, H u (ω) and H b (ω) are the system transfer functions estimated from the received signals of the upper and lower transducers, respectively.

[0082] The response of the anti-symmetric mode is calculated by equation (4):

[0083] u a (t) = ∫ E(ω) H A (ω) e iωt dω = ∫ E(ω) [H u (ω) - H b (ω)] e iωt dω (4)

[0084] where E(ω) is the Fourier transform of the response e(t) of the anti-symmetric mode.

[0085] The phase velocity dispersion curve of the guided wave is calculated from the Rayleigh-Lamb frequency equation, where the symmetric mode satisfies

[0086]

[0087] The anti-symmetric mode satisfies:

[0088]

[0089] where p and q are given by where ω is the angular frequency / rad, c L is the longitudinal wave speed, k is the wave number, and c L is the transverse wave speed. The Rayleigh-Lamb equation clearly expresses the phase velocity c pThe relationship between the frequency-plate thickness product fdalso determines that the guided wave is multi-modal and dispersive. Subsequently, the group velocity dispersion curve can be obtained by the following equation:

[0090]

[0091] Subsequently, the Hilbert transform is used to obtain the envelope curve of the narrowband response. The peak value of the envelope curve is considered as the arrival time of the selected mode. The time of flight T tof is calculated by equation (5):

[0092] T tof = T toa -T ac (5)

[0093] where the time of flight T tof is the difference between the arrival time T toa and the transmission time T ac .

[0094] According to each sensing path, the length L and the time of flight t tof , the corresponding group velocity c g can be estimated by equation (6):

[0095] c g = L / t tof (6)

[0096] The corrosion index CI n can be obtained according to equation (7):

[0097]

[0098] where t tof1 is the time of flight of the guided wave of the first working mode, t tof2 is the time of flight of the guided wave of the second working mode, L is the length from the excitation position to the receiving position, l is the length of the guided wave passing through the corrosion area, is the group velocity of the first working mode at the corrosion site, is the group velocity of the first working mode at the non-corrosion site, is the group velocity of the second working mode at the corrosion site, is the group velocity of the second working mode at the non-corrosion site.

[0099] Given the same environment and measurement conditions, the change in group velocity is caused by the thickness loss caused by corrosion. Therefore, the corrosion index can be used for corrosion detection.

[0100] To determine the corrosion location, the monitoring area is divided into a grid of uniform distribution. On this basis, the probability of each grid defect occurrence (x, y) can be estimated from the size of the group velocity variance and the position of the relative sensing path, as shown in the following formula (8):

[0101]

[0102] wherein,

[0103]

[0104] P n (x,y) is the defect distribution probability estimated from the nth sensing path. (x n1 ,y n1 ) and (x n2 ,y n2 ) are the coordinates of the excitation and reception of the sensing path, respectively. β is a scale parameter that controls the size of the elliptical distribution area. If β is too small, the algorithm is too sensitive and may introduce false results. If β is too large, the algorithm is too conservative and the resolution is not high. The value of β is usually about 1.05. The corrosion index of all sensing paths is substituted into formula (8) to obtain the reconstructed image.

[0105] The present application utilizes the high-order mode of guided waves in a large frequency-thickness product range to detect corrosion damage in metal plates. The high-order mode and wide frequency band ensure the acquisition of a large amount of alternative information, meeting the requirements of corrosion sensitivity and mode purity. To further enhance the purity of the Lamb wave mode, two completely identical sensors are symmetrically attached to the two sides of the plate as receiving sensors, and mode control is achieved by directly adding or subtracting the captured signals. In addition, a corrosion sensitivity principle is established, and the corrosion sensitivity is defined as the amount of group velocity change caused by the change of the given plate thickness, which provides a strategy for effectively determining the Lamb wave detection point. Finally, the probability diagnosis algorithm is used for imaging. The two modal signals used to represent corrosion in the present application come from the same test, and can be applied to online damage monitoring of large plate and shell structures in the fields of aviation, aerospace, energy, and ships, regardless of changes in working environment and load conditions.

[0106] The method of the present application is illustrated by the following examples.

[0107] The prototype was a 2024-T3 aluminum plate with dimensions of 1000mm × 1000mm × 3.9mm. A guided wave with a large frequency and thick layer was generated using an Innolas Spitlight 600-10YAG laser (pulse energy 190mJ, pulse width 6-7ns, wavelength 1064nm, unfocused beam diameter 6mm, repetition rate 10Hz). Two Olympus C133-RM ultrasonic probes (center frequency 2.25MHz, bandwidth [0, 4.5]MHz, diameter 6mm) served as receiving sensors. An Agilent DSOX-3014A oscilloscope was used to record the probe response signals (sampling frequency 10MHz, data length 10000 points). A mirror on the sliding base rail changed the excitation position, while the receiving position could be changed by manually moving the probes simultaneously on both sides of the plate. 0.05mm of aluminum was sanded away in a circular area on the surface with a diameter of 35mm to simulate corrosion damage.

[0108] A guided wave corrosion detection method based on high-frequency thick-layer excitation includes the following steps:

[0109] 1) Sensor installation.

[0110] like Figure 1 As shown, the detection area is planned, a coordinate system is established, and a grid is divided as the coordinates of the detection pixels. A sensor array is arranged on the part to be inspected, and the sensor position coordinates are recorded. The corrosion area is represented by a dark gray solid circle. The coordinate system composed of the horizontal x-axis and the vertical y-axis is used to describe the monitoring area, with the origin at the center of the aluminum plate. The coordinates of the center of the corrosion area are (100.7, 100.1) mm.

[0111] 2) Modal selection.

[0112] The propagation group velocity dispersion curves of guided waves of different modes in 2024-T3 aluminum plates with thicknesses of 3.9 mm and 3.4 mm are given. By comparing the velocity difference between the two plates for each mode, the mode A1, which is sensitive to thickness changes, and the mode S0, which is less sensitive to thickness changes, are selected.

[0113] 3) Select the operation point.

[0114] The accuracy of corrosion detection depends on the velocity variation of the guided wave as it propagates through the thinned region. Therefore, the chosen mode must exhibit significant dispersion. On the other hand, the measurement of the guided wave arrival time must be sufficiently accurate. However, the energy of the wave packet diffuses in the time domain due to dispersion, causing waveform changes and leading to inaccurate time-of-flight estimation. Therefore, narrowband windowed tonebursts, rather than short spike pulses, are typically chosen as the input signal. For narrowband detection, the mode and frequency of the guided wave used determine the accuracy of the corrosion detection.

[0115] Wave propagation in a plate with large product of frequency and thickness (including many modes of wide frequency band) can carry rich structural information for defect detection. From it, several narrowband wave responses can be extracted, whose frequency bands are contained in the wideband excitation frequency band. This can be an effective method to optimize guided wave detection. The present invention considers determining the appropriate operating point from the corrosion sensitivity.

[0116] The sensitivity of a guided wave mode to corrosion is defined as the amount of velocity change (increase or decrease) caused by a given change in plate thickness. In particular, the amount of velocity change when the plate thickness changes from d low to d' can be calculated as follows:

[0117] velocity variation = |c g (fd)-c g (fd')| (1)

[0118] Therefore, the sensitivity is directly related to the degree of dispersion of the selected operating point. The steeper the slope of the group velocity dispersion curve, the higher the sensitivity of the mode to the defect. Figure 2 is the group velocity dispersion curve of guided waves propagating in 2024-T3 aluminum plates. When the plate thickness decreases from 3.9 mm to 3.4 mm, the velocity change of each mode is given in Figure 3 . It can be seen that the velocity change of the S0 mode and the A0 mode caused by corrosion is not obvious in the low frequency range (≤ 0.3 MHz). In contrast, the velocity of the high-order guided wave mode changes sharply near the cutoff frequency.

[0119] Therefore, when selecting a guided wave mode, the velocity change of both the symmetric mode and the anti-symmetric mode should be calculated. By comparing the sensitivity of the two modes to corrosion, the mode with high corrosion sensitivity is selected for corrosion detection. The data obtained in this way is more accurate, and the corrosion detection is more accurate.

[0120] In practical applications, the velocity change is determined by monitoring the change in the time of flight of a wave packet from excitation to reception. If the length of the sensing path through the corrosion area is l, the amount of change in the time of flight is calculated as follows:

[0121]

[0122] Assuming that l takes the diameter of the corrosion area 35 mm, the time of flight change of each mode is calculated according to formula (2) and plotted in Figure 4 . In order to accurately collect the signal, the time of flight change Δt tofThe sampling interval must be greater than the sampling interval. In an embodiment, the sampling frequency is 10 MHz and the sampling interval is 0.1 μs. In this case, the selectable operating frequency range for the S0 mode is [0.16, 0.51] MHz, the selectable operating frequency range for the Al mode is [0.45, 0.55] MHz, [0.84, 1.17] MHz, [1.28, 1.5] MHz, the selectable operating frequency range for the S1 mode is [0.8, 0.93] MHz, [1.19, 1.5] MHz, and the selectable operating frequency range for the S2 mode is [0.83, 1.09] MHz.

[0123] By considering the corrosion sensitivity, the operating point is determined to be the Al mode at [0.46, 0.52] MHz in combination with the S0 mode at [0.36, 0.42] MHz.

[0124] 4) Calculation results.

[0125] After the selectable operating point (guided wave mode and frequency) is determined, a narrowband windowed toneburst excitation can be used to detect corrosion. The response of the symmetric mode is calculated according to equation (3):

[0126] u s (t) = ∫ E(ω) H S (ω) e iωt dω = ∫ E(ω) [H u (ω) + H b (ω)] e iωt dω (3)

[0127] where E(ω) is the Fourier transform of the symmetric mode response e(t), H u (ω) and H b (ω) are the system transfer functions estimated from the received signals of the upper and lower transducers, respectively.

[0128] The response of the anti-symmetric mode is calculated according to equation (4):

[0129] u a (t) = ∫ E(ω) H A (ω) e iωt dω = ∫ E(ω) [H u (ω) - H b (ω)] e iωt dω (4)

[0130] where E(ω) is the Fourier transform of the anti-symmetric mode response e(t).

[0131] Subsequently, the Hilbert transform is used to obtain the envelope curve of the narrowband response. The peak value of the envelope curve is considered to be the arrival time of the selected mode. The time of flight t tof is calculated from equation (5):

[0132] T tof = T toa - T ac (5)

[0133] where the time of flight T tof is the difference between the arrival time T toa and the launch time T ac .

[0134] According to the length L and the time of flight t tof of each sensing path, the corresponding group velocity c g can be estimated from equation (6):

[0135] c g = L / t tof (6)

[0136] For a defective ray, the ratio of the arrival times of the S0 mode at a non-dispersive frequency point (fl = 0.2 MHz) and at a highly dispersive frequency point (f2 = 0.57 MHz) can be expressed as:

[0137]

[0138] For a non-defective ray, / is equal to zero, and then the ratio is equal to:

[0139]

[0140] For a defective ray, the ratio of the arrival time of the Al mode at 0.47 MHz (fl) to the arrival time of the S0 mode at 0.57 MHz (f2) can be expressed as:

[0141]

[0142] For a non-defective ray, / is equal to zero, and then the ratio ratio A1S0 is equal to:

[0143]

[0144] According to equation (7), the corrosion index CI n can be obtained:

[0145]

[0146] It measures the difference in the change of velocity of two different operating modes of the sensing path after encountering a corrosion damage. Given the same environmental and measurement conditions, the change in group velocity is caused by the thickness loss due to corrosion. Therefore, the corrosion index can be used for corrosion detection.

[0147] To determine the corrosion location, the monitoring area is divided into a grid of uniformly distributed cells. Based on this, the probability of a defect occurring in each cell (x, y) can be estimated from the size of the group velocity variance and the relative position of the sensing path, as shown in equation (8):

[0148]

[0149] where,

[0150]

[0151] P n (x, y) is the probability of a defect occurring in a cell estimated from the nth sensing path. (x n1 ,y n1 ) and (x n2 ,y n2 ) are the coordinates of the excitation and receiving points of the sensing path, respectively. β is a scale parameter that controls the size of the elliptical distribution area. If β is too small, the algorithm is too sensitive and can introduce false results. If β is too large, the algorithm is too conservative and the resolution is not high. The value of β is usually around 1.05. By substituting the corrosion index of all sensing paths into equation (8), the reconstructed image can be obtained.

[0152] Fig. 5(a), Fig. 5(b), Fig. 5(c), Fig. 5(d) show the damage identification results of the mode S0 at 400 KHz and the A1 mode frequency at 460 KHz, 480 KHz, 500 KHz, 520 KHz, respectively. Fig. 5(e), Fig. 5(f), Fig. 5(g), Fig. 5(h) show the damage identification results of the mode A1 at 500 KHz and the S0 mode frequency at 360 KHz, 380 KHz, 400 KHz, 420 KHz, respectively. Each image is normalized, and it can be seen that the corrosion is correctly identified, and the identified damage location matches the actual corrosion location very well. This means that the use of the propagation characteristics of high-order guided wave modes can provide more possibilities for damage detection.

Claims

1. A method for detecting corrosion in sheet metal, characterized in that, Includes the following steps: 1) Substitute the parameters of the plate under test into the frequency equation to calculate the phase velocity and group velocity dispersion curves of the multimode guided wave. The parameters of the plate under test include: density, Young's modulus, Poisson's ratio and thickness. 2) Determine the sensitivity to corrosion damage based on the slope of the group velocity dispersion curve of each mode of guided wave. Select guided wave modes based on the sensitivity to corrosion. The sensitivity of guided wave modes to corrosion is determined by the magnitude of the change in guided wave velocity caused by the change in plate thickness. Select two guided wave modes that are highly sensitive to corrosion and have opposite trends in group velocity change under the condition of thinning of structural thickness as working modes and determine their respective working frequencies. 3) Apply broadband excitation to the board to generate multimode guided waves, and detect the guided waves at the edge of the area to be detected on the board to obtain the flight time of the two selected operating modes at their respective operating frequencies; the guided wave flight time is the difference between the guided wave arrival time and the transmission time; the guided wave arrival time is the peak value of the envelope curve, wherein the envelope curve is obtained by Hilbert transform of the narrowband response of the selected guided wave modes; the guided wave modes include symmetric modes and antisymmetric modes; 4) Calculate the ratio of the flight time of the two working modes at their respective working frequencies. This ratio is the corrosion index corresponding to the area to be detected. 5) Divide the area to be inspected into a grid, and obtain the defect distribution probability of the grid based on the positional relationship between the corresponding grid and each propagation path and the corrosion index of each propagation path.

2. The corrosion detection method for plates according to claim 1, characterized in that, In step 1), the phase velocity dispersion curve of the guided wave is calculated using the frequency equation, and the symmetric mode satisfies: ； Antisymmetric mode satisfies: ； The formula for calculating parameter p is: The formula for calculating parameter q is: , Angular frequency in rad. Let k be the longitudinal wave velocity and k be the wave number. d represents the transverse wave velocity, and d represents the constant plate thickness.

3. The corrosion detection method for plates according to claim 1, characterized in that, When selecting the guided wave mode, the change in guided wave velocity is determined by monitoring the change in the flight time required for the wave packet to propagate from the excitation position to the receiving position.

4. The corrosion detection method for plates according to claim 3, characterized in that, The formula for calculating the change in flight time is: ； Where L is the length from the excitation position to the receiving position, and l is the length of the guided wave passing through the eroded region. The group velocity of the corroded part, The group velocity is the velocity of the uncorroded portion.

5. The corrosion detection method for plates according to claim 2, characterized in that, The response of the symmetric mode for: ； in, It is a symmetric mode response. Fourier transform, and These are the system transfer functions estimated based on the received signals from the upper and lower transducers, respectively. Represents symmetric modes; The response of the antisymmetric mode for: ； in, It is an antisymmetric modal response Fourier transform, and These are the system transfer functions estimated based on the received signals from the upper and lower transducers, respectively. This indicates an antisymmetric mode.

6. The corrosion detection method for plates according to claim 1, characterized in that, In step 4), the corrosion index The calculation method is as follows: ； Among them, t tof1 The time of flight of the guided wave in the first operating mode is t. tof2 Let L be the flight time of the guided wave in the second operating mode, L be the length from the excitation position to the receiving position, and l be the length of the guided wave through the eroded region. The group velocity of the first working mode at the corroded site. The group velocity of the first working mode in the uncorroded area. The group velocity of the second working mode at the corroded site. This represents the group velocity of the second working mode in the uncorroded area.

7. The corrosion detection method for plates according to claim 1, characterized in that, In step 5), the method for calculating the relative propagation path is as follows: ； in, For the excitation position coordinates, To detect position coordinates.

8. The corrosion detection method for plates according to claim 7, characterized in that, The method for calculating the defect distribution probability of the mesh is as follows: ； in, It is a scale parameter that controls the size of the elliptical distribution region.

Images