A dynamic impedance matching method for electromagnetic acoustic testing

Abstract

The application relates to the technical field of ultrasonic detection, in particular to a dynamic impedance matching method for electromagnetic ultrasonic detection. First, a continuous mapping relationship model of complex impedance and material electromagnetic parameters and lift-off is established; before actual detection, the complex impedance of the current state of a workpiece is obtained in real time through a double-frequency detection signal, and the conductivity, the magnetic permeability and the lift-off height are calculated by inputting the model; then, the real working impedance of the coil is accurately calculated according to the parameters; the capacitance and the inductance parameters of the matching network are dynamically adjusted based on the real working impedance, so that the conjugate matching is realized; finally, the ultrasonic pulse is transmitted for detection under the optimal matching state. Through real-time sensing and adaptive adjustment, the method completely solves the mismatching problem caused by the change of the material and the state of the workpiece, ensures that the energy transmission efficiency and the echo signal-to-noise ratio are always optimal, and thus the detection accuracy and reliability are significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ultrasonic testing, and particularly relates to a dynamic impedance matching method for electromagnetic acoustic testing. BACKGROUND

[0002] Electromagnetic acoustic testing technology (EMAT) is an advanced non-contact, coupling agent-free ultrasonic nondestructive testing method. The core component of the electromagnetic acoustic transducer establishes a static bias magnetic field on the surface of the conductive workpiece through a permanent magnet, and generates a dynamic magnetic field by passing an alternating current through a high-frequency coil. Based on the Lorentz force mechanism or the magnetostrictive effect, the two are coupled to directly excite and receive ultrasonic waves in the workpiece. This method avoids the dependence of traditional piezoelectric ultrasonic on surface treatment and coupling agent, and is very suitable for high-temperature, high-speed online detection and detection of rough workpieces. High-precision thickness measurement can be achieved by measuring ultrasonic transit time.

[0003] Compared with piezoelectric transducers, the energy conversion rate of EMAT technology is one to two orders of magnitude lower, resulting in large excitation energy and low signal-to-noise ratio of received signals. In order to improve the energy transmission efficiency, the industry generally uses impedance matching technology, that is, an impedance matching network (such as an L-type or π-type network) is inserted between the excitation source (usually with a standard impedance of 50Ω) and the EMAT coil, aiming to achieve conjugate matching between the source end and the load end, thereby maximizing the transmission power and minimizing signal reflection.

[0004] However, the equivalent impedance of EMAT is not a fixed value. It is not only related to the structure of the probe itself, but also strongly depends on the electromagnetic properties (electrical conductivity, magnetic permeability) of the workpiece to be detected, the lift-off distance between the probe and the workpiece, and the operating angular frequency. For example, when detecting workpieces of different materials (such as aluminum, copper, and steel) or the same material in different heat treatment states, the electrical conductivity and magnetic permeability differ significantly, resulting in impedance mismatch. This means that the matching network optimized for a specific working condition will quickly fail when the workpiece or detection conditions change, severely affecting the stability and reliability of the detection, and therefore needs to be addressed. SUMMARY

[0005] In order to avoid and overcome the technical problems existing in the prior art, the present application provides a dynamic impedance matching method for electromagnetic acoustic testing, which dynamically and adaptively adjusts the impedance matching network to ensure maximum energy transmission efficiency and highest echo signal-to-noise ratio under any detection condition, thereby effectively improving the accuracy of thickness measurement.

[0006] To achieve the above purpose, the present application provides the following technical scheme:

[0007] A dynamic impedance matching method for electromagnetic acoustic testing, comprising the following steps:

[0008] S1, by emitting dual-frequency excitation signals to the calibration sample at different lift-off heights, simultaneously collecting the voltage and current of the electromagnetic coil to calculate the complex impedance; taking the complex impedance as the input, and the conductivity, magnetic permeability and lift-off height of the corresponding calibration sample as the label, a continuous mapping relationship model is trained;

[0009] S2, emitting dual-frequency excitation signals to the workpiece to be detected, and collecting the voltage and current of the electromagnetic coil to calculate the complex impedance, inputting the complex impedance into the mapping model to obtain the conductivity, magnetic permeability and lift-off height of the workpiece to be detected;

[0010] S3, according to the working angular frequency of the excitation circuit, the effective radius of the electromagnetic coil and the parameters obtained in step S2, the skin depth and the impedance change are calculated to obtain the working impedance of the electromagnetic coil on the surface of the workpiece to be detected;

[0011] S4, adjusting the capacitance and inductance in the impedance matching network according to the working impedance to minimize the reflection coefficient;

[0012] S5, after the impedance matching is completed, the excitation circuit emits an ultrasonic excitation pulse and receives the echo signal to measure the thickness.

[0013] As a further scheme of the application: the acquisition process of the continuous mapping relationship model is as follows:

[0014] S11, preparing calibration samples: preparing a set number of metal calibration samples whose conductivity and magnetic permeability are calibrated by the eddy current-bridge method;

[0015] S12, setting the lift-off height: for each calibration sample, a predetermined number of lift-off heights are set;

[0016] S13, signal emission and data collection: at each lift-off height, the excitation circuit sequentially injects short-period sinusoidal envelope signals of the first frequency and the second frequency into the electromagnetic coil, and in this process, the voltage U and the current I flowing through the electromagnetic coil are synchronously collected coaxially;

[0017] S14, calculating the complex impedance: according to the collected voltage and current, the complex impedance Z=U / I at two frequencies is calculated, so that a pair of complex impedances (Z1, Z2) is obtained for each calibration sample at each lift-off height, Z1 represents the complex impedance obtained under the excitation of the first frequency, and Z2 represents the complex impedance obtained under the excitation of the second frequency;

[0018] S15, taking all the complex impedance values (Z1, Z2) calculated in step S14 as input features, and taking their corresponding conductivity, magnetic permeability and lift-off height as output labels to jointly form a model training data set;

[0019] S16. Using machine learning methods, train the above training dataset to obtain a model that reflects the continuous mapping relationship from complex impedance value to conductivity, permeability and lift-off height.

[0020] As a further aspect of the present invention, the process for obtaining the conductivity, magnetic permeability, and lifting height of the workpiece to be inspected is as follows:

[0021] S21. A short-period sinusoidal envelope signal of the first frequency is emitted to the electromagnetic coil through the excitation circuit. At the same time, the voltage U1 and current I1 across the electromagnetic coil are simultaneously acquired, and the real-time complex impedance Z at this frequency is calculated. 1,t =U1 / I1;

[0022] S22. A short-period sinusoidal envelope signal of a second frequency is emitted to the electromagnetic coil through the excitation circuit. At the same time, the voltage U2 and current I2 across the electromagnetic coil are simultaneously acquired, and the real-time complex impedance Z at this frequency is calculated. 2,t =U2 / I2;

[0023] S23. Combine the two real-time complex impedances obtained in steps S21 and S22 into a real-time complex impedance pair (Z). 1,t Z 2,t ), and (Z) 1,t Z 2,t This serves as the real-time impedance characteristic input of the workpiece under inspection at the current detection point.

[0024] S24, (Z) 1,t Z 2,t The data is input into the continuous mapping relationship model, the model calculation is performed, and the conductivity, magnetic permeability and lifting height of the workpiece under inspection at the current detection point are output.

[0025] As a further aspect of the present invention, the process for obtaining the working impedance is as follows:

[0026] S31. Obtain the operating angular frequency of the excitation circuit and the effective radius of the electromagnetic coil;

[0027] S32. Calculate the skin depth using the conductivity and permeability of the workpiece to be inspected, as well as the operating angular frequency of the excitation circuit.

[0028] S33. Calculate the change in electromagnetic coil impedance caused by the presence of the workpiece by combining the skin depth with the conductivity, permeability, and lift-off height of the workpiece to be inspected, as well as the operating angular frequency of the excitation circuit and the effective radius of the electromagnetic coil.

[0029] S34. Obtain the impedance of the electromagnetic coil in free space, and add this impedance to the impedance change. The sum is the working impedance Z of the electromagnetic coil on the surface of the workpiece to be inspected.E .

[0030] As a further aspect of the present invention, the formula for calculating skin depth is as follows:

[0031] ;

[0032] In the formula, δ represents the skin depth; ω represents the operating angular frequency of the excitation circuit; σ and μ represent the electrical conductivity and magnetic permeability of the workpiece under inspection, respectively.

[0033] As a further aspect of the present invention, the formula for calculating the impedance change is as follows:

[0034] ;

[0035] In the formula, ΔZ represents the impedance change; r represents the effective radius of the electromagnetic coil; d represents the lift-off height between the electromagnetic coil and the surface of the workpiece to be inspected; e represents the natural constant; k represents the coefficient of variation; and j is the imaginary unit.

[0036] As a further aspect of the present invention: the impedance matching network is an L-type active impedance matching network, and its adjustment process is as follows:

[0037] S41, Adjust the working impedance Z E The target load impedance to be achieved by setting the L-type active impedance matching network; obtain the output impedance Z of the excitation circuit. s The output impedance Z of the L-type active impedance matching network M Z s Including the real part R S And the imaginary part X S Z M Including the real part R M And the imaginary part X M Z E Including the real part R E And the imaginary part X E ;

[0038] S42. According to the maximum power transfer theorem, in order to achieve conjugate matching between the excitation circuit and the electromagnetic coil, Z must satisfy... s =Z E , will Z s and Z E Substituting these values ​​into the capacitor and inductance calculation formulas, we can calculate the theoretical parallel capacitance C and the theoretical series inductance L. This indicates the conjugate operation on complex numbers.

[0039] S43. Compare the theoretical parallel capacitance value C and the theoretical series inductance value L with the actual capacitance and inductance element value library in the L-type active impedance matching network, select a set of capacitor and inductor combinations that are closest to the theoretical parallel capacitance value C and the theoretical series inductance value L, and make this capacitor and inductor combination in the L-type active impedance matching network in the closed circuit state, while the other capacitors and inductors are in the open circuit state. At this time, the adjustment of the L-type active impedance matching network is completed.

[0040] S44. After adjustment, the excitation circuit emits a verification signal, and the output impedance Z of the L-type active impedance matching network is measured at this time. M The reflection coefficient is calculated. If the reflection coefficient is higher than the preset threshold, the process returns to step S43, and a new pair of capacitors and inductors is selected for iterative optimization until the reflection coefficient meets the requirements.

[0041] As a further aspect of the present invention, the specific formulas for calculating capacitance and inductance are as follows:

[0042] ;

[0043] ;

[0044] In the formula, L n S represents the nth inductor connected in series in an L-type active impedance matching network; n Indicates that L n Connect the series switch in the L-type active impedance matching network; C n This represents the nth capacitor connected in parallel in the L-type active impedance matching network; Indicates that C n Connect the parallel switch in the L-type active impedance matching network; N is the total number of inductors set in the L-type active impedance matching network.

[0045] As a further aspect of the present invention, the formula for calculating the reflection coefficient is as follows:

[0046] ;

[0047] In the formula, This represents the reflection coefficient.

[0048] As a further aspect of the present invention, the thickness measurement process is as follows:

[0049] S51. An ultrasonic excitation pulse is generated by the excitation circuit and applied to the electromagnetic coil through the impedance matching network, which is now in the best matching state.

[0050] S52. The pulse current in the electromagnetic coil induces eddy currents in the workpiece to be inspected. The eddy currents interact with the bias static magnetic field to generate Lorentz force, magnetostrictive force, etc., thereby exciting ultrasonic waves in the workpiece to be inspected.

[0051] S53. When ultrasonic waves propagate in the workpiece to be inspected and encounter defects or the bottom surface, they are reflected. The returning ultrasonic vibrations cause the electrons on the surface of the workpiece to move, cutting the magnetic lines of force and inducing an echo voltage signal in the electromagnetic coil.

[0052] S54. The collected echo voltage signal is filtered and noise reduced. The propagation time of the ultrasonic wave in the workpiece to be inspected is calculated according to the ultrasonic time-of-flight principle. The thickness of the workpiece to be inspected is calculated based on the propagation time.

[0053] Compared with the prior art, the beneficial effects of the present invention are:

[0054] This invention introduces a rapid "sensing" stage before detection, utilizing dual-frequency detection signals to calculate key variables such as the workpiece's conductivity, permeability, and lift-off height in real time. Based on this, a rigorous electromagnetic theory model is used to accurately calculate the coil's true operating impedance. Finally, the matching network parameters are dynamically adjusted with conjugate matching as the target. This process represents a leap from "empirical fixed matching" to "real-time precise matching based on a physical model," ensuring that ultrasonic excitation energy is transmitted with maximum efficiency under any detection conditions, while even weak echoes are received with an extremely high signal-to-noise ratio. The ultimate result is a significant improvement in the signal-to-noise ratio of the detection signal, thereby greatly enhancing the accuracy, repeatability, and reliability of thickness measurement and defect identification. This allows the equipment to intelligently adapt to different materials and working conditions, achieving a substantial improvement in system performance. Attached Figure Description

[0055] Figure 1 This is a flowchart of the dynamic impedance matching method of the present invention.

[0056] Figure 2 This is a schematic diagram of the electromagnetic ultrasonic transducer coupling and impedance matching system with the workpiece in this invention.

[0057] Figure 3 This is a schematic diagram of the L-type impedance matching circuit of the electromagnetic ultrasonic transducer in this invention.

[0058] Figure 4 This is a schematic diagram of the L-type active impedance matching network in this invention.

[0059] Figure 5 This is a comparison diagram of the echo signals before and after impedance matching in this invention.

[0060] Figure 6 for Figure 5 A magnified view of the fourth echo signal. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] Please see Figures 1-5 In this embodiment of the invention, a dynamic impedance matching method for electromagnetic ultrasonic testing includes the following: [e.g., ...] Figure 2 The diagram illustrates the physical connection between the impedance matching device and the main unit of the electromagnetic ultrasound equipment, demonstrating the hardware composition of the system in practical applications. Figure 2 The notation Z=R+jX indicates the core function of the impedance matching device: transforming the complex impedance of EMAT (expressed as real resistance R and imaginary reactance jX). (σ,μ) emphasizes that the conductivity σ and permeability μ of the workpiece under inspection are two key material properties that determine its impedance characteristics, which is also the physical basis for achieving precise dynamic impedance matching.

[0063] Figure 3 The impedance matching circuit of an electromagnetic ultrasonic transducer is shown, consisting of an excitation circuit, an L-shaped impedance matching network, and the electromagnetic ultrasonic transducer connected sequentially. The excitation circuit includes a voltage source V. S Internal resistance R S and reactance X S The L-type impedance matching network consists of reactance X M and resistance R M The electromagnetic ultrasonic transducer is composed of reactance X E and resistance R E The L-shaped network is used to achieve impedance matching between the excitation circuit and the transducer, thereby improving energy transmission efficiency.

[0064] I. Standard Sample Calibration and Machine Learning Modeling

[0065] By transmitting dual-frequency excitation signals to the calibration sample at different lift-off heights and simultaneously acquiring the voltage and current of the electromagnetic coil, the complex impedance is calculated. Using the complex impedance as input and the conductivity, permeability, and lift-off height of the corresponding calibration sample as labels, a continuous mapping relationship model is trained.

[0066] 1. Preparation of calibration samples

[0067] Five 45# steel calibration samples were selected, and different microstructures were obtained through different heat treatment processes, resulting in different electrical conductivity σ and magnetic permeability μ.

[0068] The conductivity and permeability of each sample were precisely calibrated using the eddy current-bridge method, resulting in a set of conductivity calibration values.

[0069] σ cal =[σ1,σ2,σ3,σ4,σ5]=[6×10 6 5.6×10 6 5.2×10 6 4.4×10 6 3.6×10 6 S / m, set of permeability calibration values ​​μ cal =[μ1,μ2,μ3,μ4,μ5]=[50,80,200,300,400]×μ0, where μ0 is the free permeability, with a value of 4π×10⁻⁶. -7 H / m. σ1, σ2, σ3, σ4, σ5 and μ1, μ2, μ3, μ4, μ5 represent the electrical conductivity and magnetic permeability of the first to fifth 45# steel calibration samples, respectively.

[0070] 2. Set the lift height

[0071] A feeler gauge of known thickness was placed between the workpiece and the probe to set different lift-off heights. For each sample, five different lift-off heights were set, with values ​​d=[0,0.5,1,1.5,2]mm.

[0072] 3. Transmitted Signals and Data Acquisition

[0073] The EMAT is positioned at a specific lift-off height on a sample using a clamp. The excitation circuit sequentially transmits two short-period sinusoidal envelope signals at two frequencies: 2MHz and 500kHz, with an envelope period range of 1-5. These signals have low power and are used only for measurement; they do not generate ultrasonic waves. Simultaneously with the transmission of each frequency signal, the signal acquisition circuit synchronously and coaxially acquires the voltage U across the EMAT electromagnetic coil and the current I flowing through it at a high sampling rate.

[0074] 4. Calculate complex impedance

[0075] For each acquired voltage and current data, a Fast Fourier Transform analysis was performed to extract the fundamental component at the excitation frequency, and the complex impedance was calculated using the formula Z=U / I. A pair of complex impedances (Z1, Z2) was obtained for each sample and lift-off height combination.

[0076] 5. Construct the training dataset

[0077] A total of 25 data sets were obtained by iterating through all 5 samples and 5 lift-off heights. The complex impedance (Z1, Z2) of each data set was used as the input feature, and its corresponding known values ​​(σ, μ, d) were used as the output label to construct a multi-dimensional training dataset.

[0078] 6. Train the mapping model

[0079] A lightweight feedforward neural network was used for supervised learning on the training dataset. The input layer consisted of 4 nodes (real part of Z1, imaginary part of Z1, real part of Z2, and imaginary part of Z2), and the output layer consisted of 3 nodes (σ, μ, d). The network weights were optimized using the backpropagation algorithm, resulting in a high-precision continuous mapping model: (σ, μ, d) = f(Z1, Z2). This model was integrated into the host computer software.

[0080] II. Real-time inspection of the workpiece to be inspected

[0081] A dual-frequency excitation signal is emitted to the workpiece to be inspected, and the voltage and current of the electromagnetic coil are collected to calculate the complex impedance. The complex impedance is then input into the mapping model to obtain the conductivity, permeability, and lift-off height of the workpiece to be inspected.

[0082] 1. Calculate complex impedance

[0083] The EMAT probe is placed above the 45# steel workpiece to be inspected. Within a very short time (microseconds) before the main ultrasonic pulse is emitted, the excitation circuit sequentially emits 2MHz and 500kHz detection signals. The acquisition circuit synchronously samples the voltage (U1, U2) and current (I1, I2) twice and calculates the complex impedance Z under the two detection signals in real time. 1,t and Z 2,t .

[0084] 2. Obtain real-time conductivity, permeability, and lift-off height.

[0085] The real-time complex impedance pair (Z) 1,t Z 2,t The data is input into a pre-trained continuous mapping model, and the host computer software immediately calculates the real-time conductivity, real-time relative magnetic permeability, and real-time lifting height of the workpiece to be inspected.

[0086] III. Calculation of Working Impedance

[0087] Based on the operating angular frequency of the excitation circuit, the effective radius of the electromagnetic coil, and the parameters obtained in step S2, the skin depth and impedance change are calculated to obtain the operating impedance of the electromagnetic coil on the surface of the workpiece to be inspected.

[0088] 1. Obtain basic data

[0089] In this implementation, the operating frequency f of the excitation circuit is 5MHz, the effective radius r of the electromagnetic coil is 5mm, and the average value of the impedance of the electromagnetic coil in free space has been measured in advance, and the result is Z0=10.4+j·35.44Ω.

[0090] Simultaneously, by transmitting dual-frequency detection signals to the workpiece under inspection and inputting them into the mapping model, real-time parameters of the 45# steel workpiece have been obtained: conductivity. 7×10 6 S / m, relative permeability 200, absolute permeability 8π×1 5 H / m, lift-off height d is 0.5mm. Operating angular frequency ω=2πf, with a value of 3.14159×10⁻⁶. 7 rad / s.

[0091] 2. Calculate skin depth

[0092] Based on the basic data obtained above, the skin depth was calculated to be approximately 5.03 μm using formula (1).

[0093] (1).

[0094] 3. Calculate the impedance change.

[0095] By combining the skin depth with the conductivity, permeability and lift-off height of the workpiece to be inspected, as well as the operating angular frequency of the excitation circuit and the effective radius of the electromagnetic coil, the change in electromagnetic coil impedance caused by the presence of the workpiece to be inspected is calculated by formula (2), where the value of k ranges from 0.06 to 0.07, and in this embodiment it is 0.061. The impedance change is calculated to be approximately 23.8 + j·23.8Ω.

[0096] (2).

[0097] 4. Calculate the working impedance

[0098] Obtain the impedance of the electromagnetic coil in free space, and add this impedance to the impedance change. The sum is the working impedance Z of the electromagnetic coil on the surface of the workpiece to be inspected. E Then Z E =34.2+j·59.24Ω.

[0099] IV. Adjusting the Impedance Matching Network

[0100] Adjust the capacitors and inductors in the impedance matching network according to the working impedance to minimize the reflection coefficient.

[0101] Traditional L-type networks use fixed values ​​for inductance and capacitance, and can only be matched for a specific load impedance and frequency. For example...Figure 4 As shown, in this embodiment, the fixed L and C are replaced with an array of inductors and capacitors with multiple different values, and the switching is controlled by a switching matrix, thereby realizing the digital and programmable active adjustment of the equivalent inductance and capacitance values.

[0102] 1. Component array

[0103] Inductor array (L1-L6): Provides 6 different inductance values ​​(e.g., 1μH, 3.3μH, 6.8μH, 22μH, 47μH, 100μH), covering a wide range, and can be combined to approximate any inductance value within this range.

[0104] Capacitor array (C1-C6): Also provides 6 different capacitance values ​​(e.g., 10pF, 33pF, 100pF, 220pF, 470pF, 1000pF).

[0105] 2. Switching matrix (S1-S) 12 )

[0106] Control logic: Each inductor or capacitor is connected in series with a switch, S1-S6 control the inductor, S7-S... 12 Control capacitor.

[0107] Definitions of "1" and "0": When the switch state is "1" (ON): the switch is closed, and the corresponding inductor or capacitor is connected to the circuit. When the switch state is "0" (OFF): the switch is open, and the corresponding inductor or capacitor is not connected to the circuit and does not participate in the circuit operation.

[0108] 3. Combination method

[0109] Inductors are connected in series: when all inductor switches in the "1" state are closed, these inductors will be connected in series, and the total equivalent inductance is the sum of the inductance values ​​of all inductors in the "ON" state.

[0110] Capacitors are connected in parallel: when all the capacitors in the "1" state are closed, these capacitors will be connected in parallel, and the total equivalent capacitance is the sum of the capacitance values ​​of all the capacitors in the "ON" state.

[0111] Therefore, by controlling the "0 / 1" states of these 12 switches, 64 different inductor combinations and 64 different capacitor combinations can be generated, ultimately forming a matching network with 4096 possible impedance states, thereby achieving high-precision and wide-range impedance adjustment.

[0112] 4. Adjustment steps

[0113] The working impedance Z EThe target load impedance to be achieved by the L-type active impedance matching network is set to 34.2 + j·59.24Ω. The output impedance Z of the excitation circuit is then obtained. s (50Ω) and the output impedance Z of the L-type active impedance matching network M Z s Including the real part R S (50) and the imaginary part X S (0), Z M Including the real part R M And the imaginary part X M Z E Including the real part R E (34.2) and the imaginary part X E (59.24);

[0114] 2. According to the maximum power transfer theorem, in order to achieve conjugate matching between the excitation circuit and the electromagnetic coil, Z must satisfy... s =Z E , will Z s and Z E Substituting into the capacitor and inductance calculation formulas (3) and (4), the theoretical parallel capacitance value C (1.136nF) and the theoretical series inductance value L (2.028μH) are calculated.

[0115] (3);

[0116] (4);

[0117] S43. Compare the theoretical parallel capacitance value C and the theoretical series inductance value L with the actual capacitance and inductance component value library in the L-type active impedance matching network. The component library includes inductors: [1,3.3,6.8,22,47,100]μH; capacitors: [10,33,100,220,470,1000]pF.

[0118] Using an optimal approximation algorithm (such as finding the combination whose sum of combined values ​​has the smallest Euclidean distance from the theoretical value), select a capacitor-inductance combination that is closest to the theoretical parallel capacitor value C and the theoretical series inductance value L: Connect... =1μH, C2=33pF, C3=100pF, C6=1000pF, (series inductor, parallel capacitor), and make this capacitor-inductor combination in the L-type active impedance matching network in the closed state, while the other capacitors and inductors are in the open state. At this time, the adjustment in the L-type active impedance matching network is completed.

[0119] After adjustment, the excitation circuit emits a verification signal, and the output impedance Z of the L-type active impedance matching network is measured. MThe reflection coefficient was calculated to be 0.05 using formula (5), which is lower than the preset threshold of 0.2, indicating a successful match.

[0120] (5);

[0121] If the value exceeds the preset threshold of 0.2, a new capacitor-inductor combination needs to be selected and iteratively optimized until the reflection coefficient meets the requirements.

[0122] V. Ultrasonic Thickness Measurement

[0123] After impedance matching is completed, the excitation circuit emits ultrasonic excitation pulses and receives echo signals to measure thickness.

[0124] Once the matching is complete, the excitation circuit immediately emits a high-energy, narrow-pulse 5MHz ultrasonic excitation pulse. The pulse energy is efficiently transmitted to the EMAT coil through the matching network, which is in the optimal matching state.

[0125] EMAT excites ultrasonic waves in the workpiece to be inspected, and after receiving the reflected echo from the bottom surface, a weak voltage signal is induced in the electromagnetic coil.

[0126] The echo signal, after amplification, filtering, and digitization, is uploaded to the host computer. Automatic threshold detection or related detection algorithms are used to accurately identify the time points of the incident wave and the first bottom wave, calculating the time of flight (TOF) to be 3.12 µs. Given that the transverse wave velocity of 45# steel is approximately 3230 m / s, according to the formula: thickness = (velocity of sound)... TOF) / 2=(3230 3.12) / 2=5.04mm, the actual thickness of the workpiece to be inspected is 5mm, the calculated deviation is 0.04mm, which is within the allowable error range. It can be seen that the present invention can calculate the thickness of the workpiece relatively accurately.

[0127] By employing this invention, the signal-to-noise ratio of the detected signal is significantly improved compared to a fixed matching network. For specimens of the same thickness, the echo signal scans before and after impedance matching are as follows: Figure 5 As shown, the signal peak value increased from 35.4 to 60.4, and the signal amplitude increased by 70%, indicating that impedance matching can reduce energy loss during transmission and improve signal energy.

[0128] Figure 6The magnified comparison image of the fourth echo signal visually verifies the remarkable effectiveness of the dynamic impedance matching method of this invention. The image clearly shows that after matching using the method of this invention, the amplitude of the originally weak and diffuse fourth echo signal is significantly enhanced, the waveform becomes steeper and clearer, and the signal-to-noise ratio is greatly improved. This effect fully demonstrates that by optimizing energy transmission and reception efficiency through dynamic impedance matching, deep and weak echo signals can be effectively captured, greatly improving the accuracy and reliability of time-of-flight (TOF) measurements, and laying a solid data foundation for ultimately achieving high-precision thickness measurement and defect detection.

[0129] If the electromagnetic ultrasonic transducer model is changed or the operating angular frequency of this excitation circuit is altered at this time, it is only necessary to retest the impedance value of the transducer at the operating angular frequency, recalculate the matching element values ​​accordingly, and adjust the switching matrix so that the L-type active impedance matching network meets the matching requirements, thereby maximizing the working efficiency of the electromagnetic ultrasonic transducer and improving the detection accuracy.

[0130] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A dynamic impedance matching method for electromagnetic ultrasonic testing, characterized in that, Includes the following steps: S1. By transmitting dual-frequency excitation signals to the calibration sample at different lift-off heights and simultaneously collecting the voltage and current of the electromagnetic coil, the complex impedance is calculated. Using the complex impedance as input and the conductivity, permeability and lift-off height of the corresponding calibration sample as labels, a continuous mapping relationship model is trained. S2. A dual-frequency excitation signal is emitted to the workpiece to be inspected, and the voltage and current of the electromagnetic coil are collected to calculate the complex impedance. The complex impedance is then input into the mapping model to obtain the conductivity, permeability and lift-off height of the workpiece to be inspected. S3. Based on the operating angular frequency of the excitation circuit, the effective radius of the electromagnetic coil, and the parameters obtained in step S2, calculate the skin depth and impedance change to obtain the working impedance of the electromagnetic coil on the surface of the workpiece to be inspected. S4. Adjust the capacitor and inductor in the impedance matching network according to the working impedance to minimize the reflection coefficient; S5. After impedance matching is completed, the excitation circuit emits ultrasonic excitation pulses and receives echo signals to measure thickness.

2. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 1, characterized in that, The process of obtaining the continuous mapping relationship model is as follows: S11. Preparation of calibration samples: Prepare a set number of metal calibration samples whose conductivity and permeability are calibrated by the eddy current bridge method. S12. Set the lift-off height: Set a predetermined number of lift-off heights for each calibration sample; S13. Transmitting Signals and Data Acquisition: At each lift-off height, the excitation circuit sequentially injects short-period sinusoidal envelope signals of the first and second frequencies into the electromagnetic coil. During this process, the voltage U across the electromagnetic coil and the current I flowing through the electromagnetic coil are synchronously and coaxially acquired. S14. Calculate the complex impedance: Based on the collected voltage and current, calculate the complex impedance Z=U / I at the two frequencies respectively, thus obtaining a pair of complex impedances for each calibration sample at each lift-off height. , ), Z1 represents the complex impedance obtained under the first frequency excitation, and Z2 represents the complex impedance obtained under the second frequency excitation. S15. All complex impedance values ​​calculated in step S14 ( , The model training dataset is composed of the input features, the corresponding conductivity, permeability and lift-off height as output labels; S16. Using machine learning methods, train the above training dataset to obtain a model that reflects the continuous mapping relationship from complex impedance value to conductivity, permeability and lift-off height.

3. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 2, characterized in that, The specific process for obtaining the electrical conductivity, magnetic permeability, and lifting height of the workpiece to be inspected is as follows: S21. A short-period sinusoidal envelope signal of the first frequency is emitted to the electromagnetic coil through the excitation circuit, and the voltage across the electromagnetic coil is simultaneously acquired. With current The real-time complex impedance at that frequency was calculated. ,t =U1 / I1; S22. A short-period sinusoidal envelope signal of the second frequency is emitted to the electromagnetic coil through the excitation circuit, while the voltage across the electromagnetic coil is simultaneously acquired. With current The real-time complex impedance at that frequency was calculated. ,t =U2 / I2; S23. Combine the two real-time complex impedances obtained in steps S21 and S22 into a real-time complex impedance pair. ,t , ,t ), and ( ,t , ,t This serves as the real-time impedance characteristic input of the workpiece under inspection at the current detection point. S24, will ( ,t , ,t The data is input into the continuous mapping relationship model, the model calculation is performed, and the conductivity, magnetic permeability and lifting height of the workpiece under inspection at the current detection point are output.

4. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 3, characterized in that, The process of obtaining the working impedance is as follows: S31. Obtain the operating angular frequency of the excitation circuit and the effective radius of the electromagnetic coil; S32. Calculate the skin depth using the conductivity and permeability of the workpiece to be inspected, as well as the operating angular frequency of the excitation circuit. S33. Calculate the change in electromagnetic coil impedance caused by the presence of the workpiece by combining the skin depth with the conductivity, permeability, and lift-off height of the workpiece to be inspected, as well as the operating angular frequency of the excitation circuit and the effective radius of the electromagnetic coil. S34. Obtain the impedance of the electromagnetic coil in free space, and add this impedance to the impedance change. The sum is the working impedance Z of the electromagnetic coil on the surface of the workpiece to be inspected. E .

5. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 4, characterized in that, The formula for calculating skin depth is as follows: ； In the formula, δ represents the skin depth; ω represents the operating angular frequency of the excitation circuit; σ and μ represent the electrical conductivity and magnetic permeability of the workpiece under inspection, respectively.

6. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 5, characterized in that, The formula for calculating the change in impedance is as follows: ； In the formula, ΔZ represents the impedance change; r represents the effective radius of the electromagnetic coil; d represents the lift-off height between the electromagnetic coil and the surface of the workpiece to be inspected; e represents the natural constant; k represents the coefficient of variation; and j is the imaginary unit.

7. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 6, characterized in that, The impedance matching network is an L-type active impedance matching network, and its adjustment process is as follows: S41, Adjust the working impedance Z E The target load impedance to be achieved by setting the L-type active impedance matching network; obtain the output impedance Z of the excitation circuit. s The output impedance Z of the L-type active impedance matching network M Z s Including the real part R S And the imaginary part X S Z M Including the real part R M And the imaginary part X M Z E Including the real part R E And the imaginary part X E ; S42. According to the maximum power transfer theorem, in order to achieve conjugate matching between the excitation circuit and the electromagnetic coil, Z must satisfy... s =Z E , will Z s and Z E Substituting these values ​​into the capacitor and inductance calculation formulas, we can calculate the theoretical parallel capacitance C and the theoretical series inductance L. This indicates the conjugate operation on complex numbers; S43. Compare the theoretical parallel capacitance value C and the theoretical series inductance value L with the actual capacitance and inductance element value library in the L-type active impedance matching network, select a set of capacitor and inductor combinations that are closest to the theoretical parallel capacitance value C and the theoretical series inductance value L, and make this capacitor and inductor combination in the L-type active impedance matching network in the closed circuit state, while the other capacitors and inductors are in the open circuit state. At this time, the adjustment of the L-type active impedance matching network is completed. S44. After adjustment, the excitation circuit emits a verification signal, and the output impedance Z of the L-type active impedance matching network is measured at this time. M The reflection coefficient is calculated. If the reflection coefficient is higher than the preset threshold, the process returns to step S43, and a new pair of capacitors and inductors is selected for iterative optimization until the reflection coefficient meets the requirements.

8. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 7, characterized in that, The specific formulas for calculating capacitance and inductance are as follows: ； ； In the formula, L n This represents the nth inductor connected in series in the L-type active impedance matching network; Indicates that L n Connect the series switch in the L-type active impedance matching network; This represents the nth capacitor connected in parallel in the L-type active impedance matching network; Indicates to make Connect the parallel switch in the L-type active impedance matching network; This represents the total number of inductors set in an L-type active impedance matching network.

9. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 8, characterized in that, The formula for calculating the reflection coefficient is as follows: ； In the formula, This represents the reflection coefficient.

10. The dynamic impedance matching method for electromagnetic ultrasonic testing according to claim 9, characterized in that, The thickness measurement process is as follows: S51. An ultrasonic excitation pulse is generated by the excitation circuit and applied to the electromagnetic coil through the impedance matching network, which is now in the best matching state. S52. The pulse current in the electromagnetic coil induces eddy currents in the workpiece to be inspected. The eddy currents interact with the bias static magnetic field to generate Lorentz force, thereby exciting ultrasonic waves in the workpiece to be inspected. S53. When ultrasonic waves propagate in the workpiece to be inspected and encounter defects or the bottom surface, they are reflected. The returning ultrasonic vibrations cause the electrons on the surface of the workpiece to move, cutting the magnetic lines of force and inducing an echo voltage signal in the electromagnetic coil. S54. The collected echo voltage signal is filtered and noise reduced. The propagation time of the ultrasonic wave in the workpiece to be inspected is calculated according to the ultrasonic time-of-flight principle. The thickness of the workpiece to be inspected is calculated based on the propagation time.

Images