A material micro-area nondestructive testing system and method based on physical field perturbation

Abstract

This invention discloses a non-destructive testing system and method for micro-areas of materials based on physical field perturbation. The system includes an electrolytic cell, a three-electrode system, an ultrasonic excitation module, an electrochemical potentiostatic / current-constant module, a synchronous triggering and data acquisition module, and a signal processing unit. The electrolytic cell contains an electrolyte solution and fixes the workpiece under test. In the three-electrode system, the working electrode is electrically connected to the workpiece under test. The ultrasonic excitation module emits focused ultrasonic waves with specific parameters, the electrochemical module applies a potential or scans, the synchronous acquisition module records the ultrasonic emission time and the electrochemical response signal, and the signal processing unit outputs the defect category and residual stress value through a machine learning model. This solution solves the problem that existing technologies cannot perform synchronous in-situ quantitative detection, achieving accurate characterization of micro-defects and residual stress in an electrolyte environment.

Description

Technical Field

[0001] This invention relates to the field of nondestructive testing technology for materials, and in particular to a nondestructive testing system and method for micro-area materials based on physical field perturbation. Background Technology

[0002] During service in an electrolyte environment, metal components are prone to micro-defects such as microcracks and pitting corrosion, accompanied by residual stress accumulation, directly affecting the component's service safety and lifespan. In existing technologies, electrochemical impedance spectroscopy is insensitive to micron-sized closed cracks and early pitting corrosion, easily leading to missed detections; ultrasonic testing struggles to capture signals from microcracks with openings smaller than 1 μm and cannot obtain information on the electrochemical activity of defects; residual stress testing methods often require removal from the service environment, preventing in-situ monitoring, and severe signal attenuation in liquid coupling agents results in poor repeatability. In summary, existing technologies cannot simultaneously achieve highly sensitive, quantitative, and non-destructive assessment of micro-defect morphology and residual stress levels under real electrolyte conditions, failing to meet the demands of precise component condition monitoring in industrial settings; a technical solution that can simultaneously address these issues is urgently needed. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a material micro-area non-destructive testing system based on physical field perturbation. The system includes an electrolytic cell, a three-electrode system, an ultrasonic excitation module, an electrochemical potentiostatic / current-constant module, a synchronous triggering and data acquisition module, and a signal processing unit. The electrolytic cell contains an electrolyte solution and fixes the workpiece under test. The three-electrode system includes at least a working electrode, a reference electrode, and a counter electrode, with the working electrode electrically connected to the workpiece. The ultrasonic excitation module emits focused ultrasonic waves to the workpiece at a frequency of [insert frequency here]. to Pulse width is to Peak power is to Focal spot diameter The electrochemical potentiostat / galiostat module applies a constant potential or performs a potentiodynamic scan on the three-electrode system. The constant potential ranges relative to the standard hydrogen electrode. to The electrodynamic scanning rate is to The synchronous triggering and data acquisition modules record the ultrasonic emission time under the same clock reference. and electrochemical response signals Electrochemical response signal At least including microcurrent transients Rate of change of charge transfer resistance and the rate of change of double-layer capacitance The signal processing unit is based on the electrochemical response signal. The machine learning model is used to output the defect category and residual stress value.

[0004] Preferably, the working electrode lead of the three-electrode system is a coaxial shielded wire. The coaxial shielded wire is connected to the workpiece by spot welding near the workpiece under test, and the connection part is insulated and encapsulated. The shielding layer of the coaxial shielded wire is grounded at a single point on the outer shell of the electrolytic cell. All electrical connection points that may be exposed to the electrolyte environment are coated with silicone.

[0005] Further preferably, the metal casing of the ultrasonic excitation module is grounded, and the transducer of the ultrasonic excitation module integrates a PTC thermistor. When the transducer temperature exceeds... At this time, the PTC thermistor limits the output power of the ultrasonic excitation module through the drive circuit; the ultrasonic parameters of the ultrasonic excitation module adopt any of the following combinations: Combination 1 is frequency Pulse width Peak power Combination 2 is frequency Pulse width Peak power Combination 3 is frequency Pulse width Peak power .

[0006] In a further preferred embodiment, the synchronous triggering and data acquisition module uses an FPGA-based board as the core timing controller. After receiving the start command from the host computer, the FPGA board generates two TTL pulse signals from the same clock source. The pulse width of the TTL pulse signals is... One TTL pulse signal drives the pulse generator of the ultrasonic excitation module, while the other TTL pulse signal serves as the external trigger signal for the high-speed data acquisition unit. The sampling rate of the high-speed data acquisition unit... The resolution is The start-time deviation between ultrasonic emission and electrochemical signal acquisition .

[0007] In a further preferred embodiment, the electrolytic cell adopts a double-layer container structure, with the inner layer made of PMMA and the outer layer made of transparent polycarbonate. An optical liquid leakage sensor is installed in the sandwich of the double-layer structure. When the optical liquid leakage sensor detects electrolyte leakage, it triggers an audible and visual alarm and cuts off the system power supply. The electrolytic cell is equipped with a micro-circulation chamber for electrolyte controlled by a micro-pump and valves.

[0008] Further preferably, the machine learning model of the signal processing unit is a classification-regression joint model. The core architecture of the model consists of a one-dimensional convolutional neural network and a long short-term memory network. The one-dimensional convolutional neural network extracts local features from the electrochemical transient response waveform, while the long short-term memory network captures the dependence of the signal on time evolution. The model outputs the defect category and residual stress value in parallel through a fully connected layer. The defect category is pitting, microcracks, or stress concentration area, and the residual stress value ranges from [value missing]. to .

[0009] A nondestructive testing method for micro-areas of materials based on physical field perturbation, applied to any of the above-described nondestructive testing systems for micro-areas of materials based on physical field perturbation, includes the following steps:

[0010] S1: Place the workpiece to be tested in an electrolytic cell and establish a three-electrode system consisting of a working electrode, a reference electrode, and a counter electrode. The working electrode is electrically connected to the workpiece to be tested.

[0011] S2: A constant potential is applied to the three-electrode system or a potentiodynamic scan is performed using an electrochemical potentiostatic / current galvanostatic module. The constant potential ranges relative to the standard hydrogen electrode. to The electrodynamic scanning rate is to Simultaneously, the initial electrochemical impedance spectroscopy of the workpiece under test was acquired. ;

[0012] S3: Trigger the ultrasonic excitation module. The ultrasonic excitation module emits focused ultrasonic waves towards the workpiece under test, causing periodic elastic deformation on the surface or inside of the workpiece. The frequency of the ultrasonic waves is... to Pulse width is to Peak power is to Focal spot diameter ;

[0013] S4: Electrochemical response signals are acquired within the ultrasonic emission window via a synchronous triggering and data acquisition module. Electrochemical response signal At least including microcurrent transients Rate of change of charge transfer resistance and the rate of change of double-layer capacitance ;

[0014] S5: Calculate the change in impedance spectrum under conditions with and without ultrasonic disturbance. From the electrochemical response signal and impedance spectrum change Extracting feature vectors ;

[0015] S6: Transfer the feature vector Input a pre-trained classification-regression joint model and output the defect category and residual stress value.

[0016] More preferably, after applying a constant potential in step S2, the current drift must be satisfied. And continue Under the given conditions, initial electrochemical impedance spectroscopy was then acquired. When performing potentiodynamic scanning, the current sampling frequency is set to... .

[0017] More preferably, the ultrasonic dwell time of the ultrasonic excitation module in step S3 is set to... It works in conjunction with a robotic arm to scan the workpiece being measured. The robotic arm's step distance is... Monitoring during the scanning process and The fluctuation range, when the fluctuation range is all less than The system determines the signal to be valid when the current exceeds the range. If the current exceeds the range, the system will interrupt the scan and prompt the user to check the coupling status.

[0018] Further preferred, in step S5 Calculated by fitting electrochemical impedance spectroscopy. Obtained analytically through equivalent circuit model. For microcurrent transients The rate of change; eigenvectors It also includes the transient peak change rate. Relaxation time signal rise time and the spectral energy of specific frequency bands Relaxation time The fitting equation is obtained by solving the transient waveform of the microcurrent using a double exponential decay fitting method. , Weighted average relaxation time Spectral energy of a specific frequency band By analyzing the electrochemical response signal The integral equation is obtained by performing a Fourier transform and integrating within the characteristic frequency band. ,in , , This represents the amplitude of the spectrum after Fourier transform.

[0019] Technical Effects: This invention innovatively employs electrochemical-ultrasonic coupling technology, combined with synchronous triggering acquisition and a machine learning model, to solve the core problem of existing technologies' inability to synchronously and quantitatively detect micro-defects and residual stress in an electrolyte environment. The ultrasonic excitation module generates focused ultrasonic waves with specific parameters to activate the electrochemical response of micro-defects, while the synchronous acquisition module accurately captures the response signal. The machine learning model enables the synchronous output of defect categories and residual stress, achieving high-sensitivity and quantitative characterization of micro-defects and residual stress under electrolyte conditions. Attached Figure Description

[0020] Figure 1 This is a connection block diagram of the material micro-area non-destructive testing system based on physical field perturbation in this application;

[0021] Figure 2 This is a flowchart of the material micro-area non-destructive testing method based on physical field disturbance in this application. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0023] Existing technologies cannot simultaneously and quantitatively detect micro-defects and residual stress in metal components in an electrolyte environment, resulting in incomplete detection and insufficient accuracy.

[0024] Based on this, this embodiment provides a material micro-area non-destructive testing system based on physical field perturbation, including an electrolytic cell, a three-electrode system, an ultrasonic excitation module, an electrochemical potentiostatic and current-constant module, a synchronous triggering and data acquisition module, and a signal processing unit; the electrolytic cell is used to contain the electrolyte solution and fix the workpiece under test, and the electrolyte solution can be configured according to the service environment of the workpiece under test, such as simulating seawater content. Or industrial electrolyte containing The workpiece under test is fixed by a special fixture to ensure that the detection area is completely immersed in the electrolyte and has good conductive contact with the working electrode. The fixture is made of corrosion-resistant and insulating material to avoid interference with the electrochemical detection circuit. The three-electrode system includes at least a working electrode, a reference electrode, and a counter electrode. The working electrode is made of conductive material compatible with the material of the workpiece under test, such as a platinum electrode or a graphite electrode, to ensure that no chemical reaction occurs during the detection process and interferes with the detection results. The reference electrode is a saturated calomel electrode or a silver-silver chloride electrode to provide a stable potential reference. The counter electrode is an inert electrode such as a platinum mesh to ensure stable electrolytic reaction. The working electrode and the workpiece under test are electrically connected by wire welding or clamping with a conductive fixture, and the contact resistance is controlled within a certain range. Within this range, the influence of contact resistance on the electrochemical signal is reduced; the ultrasonic excitation module emits focused ultrasonic waves towards the workpiece under test, and the ultrasonic frequency is set to [value missing]. to Among them, low frequency to Suitable for penetration testing of thick-walled components, high frequency to Suitable for detecting defects in thin-walled components or near-surface surfaces, pulse width to To ensure ultrasonic temporal resolution and avoid signal superposition from defects at different locations, peak power is required. to Adjustments are made based on the detection distance and component thickness to ensure sufficient periodic elastic deformation to activate the electrochemical response of micro-defects, and the focal spot diameter. This enables micro-area localization detection, avoiding signal confusion caused by large-area disturbances; the electrochemical potentiostatic and galvanostatic module applies a constant potential to the three-electrode system or performs potentiodynamic scanning, with the constant potential ranging from the standard hydrogen electrode value. to It covers the stable potential range of most metal components in an electrolyte environment, avoiding overpotential-induced additional corrosion or passivation of components, and has a potentiodynamic scanning rate. to slow scan rate to Used for precise capture of changes in electrochemical impedance, with a fast scan rate. to Used for rapid screening; the synchronous triggering and data acquisition modules record the ultrasound emission time under the same clock reference. and electrochemical response signals The clock reference uses a high-precision crystal oscillator, ensuring frequency stability. To ensure timing control accuracy and electrochemical response signal At least including microcurrent transients Rate of change of charge transfer resistance and the rate of change of double-layer capacitance Microcurrent transients Acquired via a high-precision current amplifier, range to accuracy It can capture weak current changes and the rate of change of charge transfer resistance. and the rate of change of double-layer capacitance The electrochemical impedance spectroscopy was obtained by fitting the Randle equivalent circuit, and the least squares method was used in the fitting process to ensure accuracy; the signal processing unit was based on the electrochemical response signal. The machine learning model outputs defect categories and residual stress values. Before receiving the signal, the machine learning model needs to undergo low-pass filtering and normalization preprocessing. The low-pass filter cutoff frequency is set to... High-frequency noise is removed, and normalization scaling is applied to scale the signal amplitude to [value missing]. to This solution ensures data quality consistency across a given range. Through multi-module collaboration, an electrolytic cell provides a stable electrolyte environment, a three-electrode system constructs a reliable electrochemical detection circuit, an ultrasonic excitation module activates the electrochemical response of micro-defects by disturbing the material with ultrasonic waves using specific parameters, an electrochemical module applies potential or scans to establish detection conditions, a synchronous acquisition module ensures signal timing consistency, and a signal processing unit achieves precise output through a model, overcoming the limitations of existing detection technologies.

[0025] The lead connections of the three-electrode system are susceptible to electromagnetic interference and corrosion, which affects signal stability.

[0026] Based on this, the working electrode lead of the three-electrode system adopts characteristic impedance. The core wire is made of oxygen-free copper, and the insulation layer is polytetrafluoroethylene (PTFE). The oxygen-free copper core wire ensures excellent conductivity, and the PTFE insulation layer has corrosion resistance and high temperature resistance. The coaxial shielded wire carries welding current near the workpiece being tested. to Welding time to The spot welding process is used to connect the device to the workpiece under test, ensuring connection strength and conductivity stability. The connection area is insulated and encapsulated with E-51 epoxy resin, with a coating thickness of [missing information]. to The welding area is completely covered to prevent corrosion caused by electrolyte contact; the shielding layer of the coaxial shielding cable is grounded at a single point on the grounding terminal of the electrolytic cell shell, and the grounding resistance is controlled within a certain range. Within this range, high-frequency electromagnetic radiation interference generated during ultrasonic transducer operation is effectively suppressed; all electrical connection points that may be exposed to the electrolyte environment are coated with food-grade corrosion-resistant silicone such as Dow Corning 734, with a coating thickness of [missing information]. to Ensure complete sealing to prevent electrolyte corrosion from causing connection failure. Coaxial shielding and single-point grounding effectively isolate electromagnetic interference, spot welding ensures conductivity stability, and insulating encapsulation and silicone coating prevent electrolyte corrosion, ensuring long-term stable operation of the three-electrode system and guaranteeing the accuracy of electrochemical signal acquisition.

[0027] The ultrasonic excitation module is prone to overheating during operation, and different scenarios have different requirements for ultrasonic parameters, which affects the applicability and safety of the detection.

[0028] Based on this, the metal casing of the ultrasonic excitation module adopts a cross-sectional area of The copper grounding wire is connected to the earth, and the grounding resistance is... Enhance electrical safety; the transducer of the ultrasonic excitation module integrates a Curie temperature of... Positive temperature coefficient thermistors, when the transducer temperature exceeds At this time, the resistance of the thermistor increases sharply, and the output power of the ultrasonic excitation module is automatically limited by the drive circuit. Within this range, overheat protection is implemented; the ultrasonic parameters of the ultrasonic excitation module adopt any combination of the following: Combination 1 is frequency Pulse width Peak power Suitable for high-precision scanning of laboratory flat test blocks, high frequency ensures axial resolution, narrow pulse reduces signal interference, and medium power avoids workpiece damage; combination two is frequency Pulse width Peak power Suitable for near-surface pitting detection, high frequency improves near-surface resolution, low power prevents pit expansion; combined with frequency. Pulse width Peak power Suitable for inspecting thick-walled industrial components, low frequency ensures penetration capability, high power ensures elastic deformation amplitude, and wide pulse enhances signal strength. PTC thermistors provide overheat protection, grounding enhances safety, and multiple parameter sets are suitable for different testing scenarios, improving system applicability.

[0029] Timing deviations between synchronous triggering and data acquisition can lead to signal mismatch and affect detection accuracy.

[0030] Based on this, the synchronization triggering and data acquisition module uses a Xilinx Artix-7 series FPGA board as the core timing controller. The FPGA board firmware is written in Verilog HDL, resulting in low logic latency and precise timing control. After receiving the start command from the host computer, it... A high-precision clock source generates two TTL pulse signals, the pulse width of which is... Rising time To ensure a steep and distortion-free trigger signal, one TTL pulse signal drives the pulse generator of the ultrasonic excitation module, controlling the timing of ultrasonic wave emission. The other TTL pulse signal serves as the external trigger signal for the high-speed data acquisition unit, which uses the ADIAD7606 chip with a sampling rate of [missing information]. The resolution is It can quickly and accurately acquire weak electrochemical response signals; through precise control of the timing logic inside the FPGA, the start time deviation between ultrasonic emission and electrochemical signal acquisition is minimized. This ensures signal timing synchronization. The FPGA board guarantees timing control accuracy, two synchronous TTL pulse signals ensure precise synchronization between ultrasonic emission and signal acquisition, and a high-speed, high-resolution acquisition unit fully captures weak electrochemical response signals, reducing signal distortion and improving data reliability.

[0031] Electrolytic cells are prone to leakage, which affects the safety and stability of testing.

[0032] Based on this, the electrolytic cell adopts a double-layer container structure, with the inner layer made of PMMA with a thickness of [missing information]. to The PMMA material is transparent, making it easy to observe the testing area. The outer layer is made of transparent polycarbonate with a thickness of [missing information]. to High-strength polycarbonate prevents impact damage; the spacing between the two layers... to A sealed interlayer is formed; an infrared photoelectric sensor is installed in the interlayer of the double-layer structure to emit wavelengths. Receiver sensitivity An optical liquid leak sensor is installed at the bottom of the interlayer. When it detects electrolyte leakage into the interlayer, it immediately triggers an acoustic intensity sensor. Light wavelength The audible and visual alarm is activated via a relay. Internal power cut-off to prevent leakage from escalating; electrolytic cells are equipped with flow range settings. to Pressure range to Peristaltic pump and response time The electrolyte microcirculation chamber is controlled by a solenoid valve, with a volume of to This ensures the electrolyte circulates smoothly, maintaining uniform concentration and temperature to prevent localized high concentrations from affecting test results. The dual-layer structure and leak sensor provide leak protection and alarm functionality, while the micro-pump and valve-controlled micro-circulation chamber guarantee a stable electrolyte supply, avoiding leakage risks and ensuring long-term safe operation of the system.

[0033] The signal processing unit needs to accurately classify defects and output quantitative residual stress, and existing models cannot achieve the same level of accuracy in both areas.

[0034] Based on this, the machine learning model of the signal processing unit is a classification-regression joint model. The core architecture of the model consists of a one-dimensional convolutional neural network and a long short-term memory network; the one-dimensional convolutional neural network structure is input layer → convolutional layer 1 (number of convolutional kernels) kernel size Step length → Pooling layer 1 (pooling kernel size) Step length → Convolutional layer 2 (number of convolutional kernels) kernel size Step length → Pooling layer 2 (pooling kernel size) Step length → Fully connected layer 1 (number of neurons) The convolutional layers extract local features from the electrochemical transient response waveform through a sliding window, such as peak changes and waveform inflection points. The pooling layers reduce data dimensionality while retaining key features. The long short-term memory network structure is input layer → LSTM layer 1 (number of hidden units). dropout rate → LSTM layer 2 (number of hidden units) dropout rate → Fully connected layer 2 (number of neurons) The LSTM layer uses a gating mechanism to capture the dependence of the signal on time evolution, and the dropout rate prevents the model from overfitting. The model outputs defect categories and residual stress values ​​in parallel through fully connected layers. The defect categories are output through a softmax activation function and are divided into pitting, microcracks, or stress concentration areas. Pitting corresponds to a feature response with a large rate of change in the double-layer capacitance, microcracks correspond to a feature response with a significant rate of change in charge transfer resistance and transient microcurrent, and stress concentration areas correspond to a feature response with a significant change in the Euclidean distance of the feature vector. The residual stress value is output through a linear activation function, with a value range of [value missing]. to This method covers the residual stress range of most engineering metal components. A one-dimensional convolutional neural network and a long short-term memory network work together to extract signal features, and a classification-regression joint model outputs results in parallel, enabling accurate identification of defect types and quantitative calculation of residual stress.

[0035] Existing detection methods cannot achieve simultaneous detection of micro-defects and residual stress in an electrolyte environment through standardized procedures, and the operational logic is unclear.

[0036] Based on this, this embodiment provides a non-destructive testing method for micro-regions of materials based on physical field perturbation, including the following steps:

[0037] S1: Place the workpiece to be tested in an electrolytic cell and establish a three-electrode system consisting of a working electrode, a reference electrode, and a counter electrode. The working electrode is electrically connected to the workpiece. The installation position of the three-electrode system ensures that the three electrodes are evenly distributed around the detection area of ​​the workpiece, and the spacing is controlled within [missing information]. to Contact resistance controlled at Within this range, ensure a uniform electric field distribution;

[0038] S2: A constant potential is applied to the three-electrode system or a potentiodynamic scan is performed using an electrochemical potentiostatic and galvanostatic module. The range of the constant potential relative to the standard hydrogen electrode is: to This is achieved using an electrochemical potentiostat such as the CHI660E, with high potential output accuracy. The electrodynamic scanning rate is to The onset and termination potentials are determined based on the polarization curve of the material being tested to avoid exceeding the stable potential range. Simultaneously, the initial electrochemical impedance spectroscopy of the workpiece is acquired. sampling frequency range to Number of collection points to Points to ensure coverage of the critical frequency range of impedance response;

[0039] S3: Trigger the ultrasonic excitation module. The ultrasonic excitation module emits focused ultrasonic waves towards the workpiece under test, causing periodic elastic deformation on the surface or inside of the workpiece. The ultrasonic frequency is... to Pulse width is to Peak power is to Focal spot diameter The emission direction is adjusted according to the detection plan. For contact testing, the ultrasonic wave is incident perpendicularly on the workpiece surface. For immersion testing, the angle is adjusted to avoid reflection interference. The periodic elastic deformation frequency is consistent with the ultrasonic wave frequency, and the deformation amplitude... To prevent damage to the workpiece;

[0040] S4: Electrochemical response signals are acquired within the ultrasonic emission window via a synchronous triggering and data acquisition module. The ultrasonic emission window time range is the time range after the ultrasonic wave is emitted. to The electrochemical response signal covers the entire process from the generation to the decay of elastic deformation. At least including microcurrent transients Rate of change of charge transfer resistance and the rate of change of double-layer capacitance Data is acquired via a high-speed data acquisition card, with a sampling interval of [missing information]. This ensures that transient signal details are captured.

[0041] S5: Calculate the change in impedance spectrum under conditions with and without ultrasonic disturbance. This calculation involves subtracting complex numbers at each frequency point, subtracting the real and imaginary parts separately to obtain the amplitude and phase of the impedance change, from the electrochemical response signal. and impedance spectrum change Extracting feature vectors eigenvectors include , , , , Each parameter was calculated using MATLAB software, and the calculation process employed numerical methods to ensure accuracy.

[0042] S6: Transfer the feature vector Input a pre-trained classification-regression joint model, output defect category and residual stress value. The model is embedded in a processor such as ARM Cortex-A9, and the output response time is [not specified]. This method meets the needs of real-time detection. Following a logical flow of establishing a system, applying a potential, acquiring a baseline, acquiring ultrasonic excitation signals, extracting features, and outputting the model, the detection objective is achieved step by step. Each step has clearly defined parameters and repeatable operations, ensuring a standardized and orderly detection process.

[0043] The instability of the electrochemical state after applying a potential can lead to distortion of the initial impedance spectrum, affecting subsequent comparative analysis. Therefore, after applying a constant potential in step S2, current drift must be satisfied. And continue The conditions are monitored in real time by an electrochemical potentiostat, with sampling intervals... ,continuous If the current change in each sample meets the requirements, the system is considered stable. Once the current drift stabilizes and meets the requirements, the initial electrochemical impedance spectroscopy is automatically triggered. Data acquisition; during potentiodynamic scanning, the current sampling frequency is set to... This is achieved through parameter settings on a high-speed data acquisition card, ensuring the capture of detailed changes in microcurrent transients. This provides reliable baseline data for subsequent impedance spectrum change calculations and feature extraction, thereby improving detection accuracy.

[0044] If the ultrasonic dwell time and step spacing are unreasonable during the scanning process, or if the signal validity cannot be determined, it will affect the detection efficiency and data reliability.

[0045] Based on this, the ultrasonic dwell time of the ultrasonic excitation module in step S3 is set to... This ensures that ultrasonic excitation at each detection point fully generates elastic deformation and that the electrochemical response signal is stably acquired, in conjunction with the step angle. .

[0046] Subdivision coefficient A stepper motor-controlled robotic arm scans the workpiece being measured, achieving high positioning accuracy. The step spacing is Ensure that the entire testing area is covered without any omissions during the scanning process; monitor in real time during the scanning process. and The fluctuation range is calculated after each data point is collected. When the fluctuation range is less than [a certain value], the fluctuation value is determined. The signal is deemed valid and the data is stored. If the current exceeds the range (i.e., the set range is exceeded), the signal is checked. When interrupted, the system interrupts the scan via a program interrupt and displays a prompt to check the coupling status, such as whether the coupling agent between the ultrasonic probe and the workpiece is sufficient or whether the electrochemical electrode connection is good. Reasonable dwell time and step spacing balance detection efficiency and resolution, while signal validity criteria and anomaly handling mechanisms ensure the quality of the acquired data.

[0047] Incomplete feature vector extraction can affect the accuracy of model output, so key feature parameters need to be supplemented.

[0048] Based on this, in step S5 The equivalent circuit was obtained by fitting the Randles circuit using the least squares method. The equivalent circuit includes the solution resistance. Charge transfer resistance and double-layer capacitor Fitting error , The parameters are obtained through synchronous calculation of the fitting parameters of the equivalent circuit. For microcurrent transients In time interval internal change and The ratio reflects the rate of change of current; eigenvector It also includes the transient peak change rate. Relaxation time signal rise time and the spectral energy of specific frequency bands ,in The peak value of the transient microcurrent and the steady current before ultrasonic excitation The ratio reflects the magnitude of the current change. The fitting equation is obtained by solving the transient waveform of the microcurrent using a double exponential decay fitting method. ,in and These are the amplitudes of the double exponential decay components, respectively. and These are the relaxation times for the corresponding components. Weighted average relaxation time This reflects the signal attenuation characteristics. For microcurrent from peak Rising to peak The time required reflects the signal rise rate. By analyzing the electrochemical response signal conduct After obtaining the frequency spectrum using the point fast Fourier transform, to The integral equation is obtained by integrating within the characteristic frequency band. ,in , , The spectral amplitude after Fourier transform reflects the energy distribution of the characteristic frequency band. Multi-dimensional characteristic parameters comprehensively characterize the electrochemical and ultrasonic perturbation response properties of the material, providing rich input information for machine learning models and further improving the accuracy of defect classification and residual stress calculation.

[0049] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A material micro-area non-destructive testing system based on physical field perturbation, characterized in that, The device comprises an electrolytic cell, a three-electrode system, an ultrasonic excitation module, an electrochemical constant potential / current module, a synchronous trigger and data acquisition module, and a signal processing unit; the electrolytic cell is used for containing an electrolyte solution and fixing a workpiece to be measured; the three-electrode system comprises at least a working electrode, a reference electrode and a counter electrode, and the working electrode is in conductive connection with the workpiece to be measured; the ultrasonic excitation module emits focused ultrasonic waves to the workpiece to be measured, the frequency of the ultrasonic waves is to , the pulse width is to , the peak power is to , and the focal spot diameter is ; the electrochemical constant potential / current module applies a constant potential to the three-electrode system or carries out a dynamic potential scanning, the value range of the constant potential relative to the standard hydrogen electrode is to , and the dynamic potential scanning rate is to . The synchronous triggering and data acquisition modules record the ultrasonic emission time under the same clock reference. and electrochemical response signals Electrochemical response signal At least including microcurrent transients Rate of change of charge transfer resistance and the rate of change of double-layer capacitance The signal processing unit is based on the electrochemical response signal. The machine learning model is used to output the defect category and residual stress value.

2. The material micro-area nondestructive testing system based on physical field perturbation according to claim 1, characterized in that, The working electrode lead of the three-electrode system uses a coaxial shielded wire. The coaxial shielded wire is connected to the workpiece by spot welding near the workpiece under test, and the connection part is insulated and encapsulated. The shielding layer of the coaxial shielded wire is grounded at a single point on the outer shell of the electrolytic cell. All electrical connection points that may be exposed to the electrolyte environment are coated with silicone.

3. The material micro-area non-destructive testing system based on physical field perturbation according to claim 1, characterized in that, The metal casing of the ultrasonic excitation module is grounded, and the transducer of the ultrasonic excitation module integrates a PTC thermistor. When the transducer temperature exceeds... At this time, the PTC thermistor limits the output power of the ultrasonic excitation module through the drive circuit; the ultrasonic parameters of the ultrasonic excitation module adopt any of the following combinations: Combination 1 is frequency Pulse width Peak power Combination 2 is frequency Pulse width Peak power Combination 3 is frequency Pulse width Peak power .

4. The material micro-area non-destructive testing system based on physical field perturbation according to claim 1, characterized in that, The synchronous triggering and data acquisition module uses an FPGA-based board as the core timing controller. After receiving the start command from the host computer, the FPGA board generates two TTL pulse signals from the same clock source. The pulse width of the TTL pulse signals is... One TTL pulse signal drives the pulse generator of the ultrasonic excitation module, while the other TTL pulse signal serves as the external trigger signal for the high-speed data acquisition unit. The sampling rate of the high-speed data acquisition unit... The resolution is The start-time deviation between ultrasonic emission and electrochemical signal acquisition .

5. The material micro-area nondestructive testing system based on physical field perturbation according to claim 1, characterized in that, The electrolytic cell adopts a double-layer container structure, with the inner layer made of PMMA and the outer layer made of transparent polycarbonate. An optical liquid leakage sensor is installed in the sandwich of the double-layer structure. When the optical liquid leakage sensor detects electrolyte leakage, it triggers an audible and visual alarm and cuts off the system power supply. The electrolytic cell is equipped with a micro-circulation chamber for electrolyte controlled by a micro-pump and valves.

6. The material micro-area nondestructive testing system based on physical field perturbation according to claim 1, characterized in that, The machine learning model of the signal processing unit is a classification-regression joint model. The core architecture of the model consists of a one-dimensional convolutional neural network and a long short-term memory network. The one-dimensional convolutional neural network extracts local features in the electrochemical transient response waveform, and the long short-term memory network captures the dependence of the signal on the evolution of time. The model outputs the defect category and residual stress value in parallel through a fully connected layer. The defect categories are pitting, microcracks, or stress concentration zones, and the residual stress values ​​range from [value missing]. to .

7. A method for nondestructive testing of micro-areas of materials based on physical field perturbation, applied to the nondestructive testing system for micro-areas of materials based on physical field perturbation as described in any one of claims 1-6, characterized in that, Includes the following steps: S1: Place the workpiece to be tested in an electrolytic cell and establish a three-electrode system consisting of a working electrode, a reference electrode, and a counter electrode. The working electrode is electrically connected to the workpiece to be tested. S2: A constant potential is applied to the three-electrode system or a potentiodynamic scan is performed using an electrochemical potentiostat / current galvanostat module. The constant potential ranges relative to the standard hydrogen electrode as follows: to The electrodynamic scanning rate is to Simultaneously, the initial electrochemical impedance spectroscopy of the workpiece under test was acquired. ; S3: Trigger the ultrasonic excitation module. The ultrasonic excitation module emits focused ultrasonic waves towards the workpiece under test, causing periodic elastic deformation on the surface or inside of the workpiece. The frequency of the ultrasonic waves is... to Pulse width is to Peak power is to Focal spot diameter ; S4: Electrochemical response signals are acquired within the ultrasonic emission window via a synchronous triggering and data acquisition module. Electrochemical response signal At least including microcurrent transients Rate of change of charge transfer resistance and the rate of change of double-layer capacitance ; S5: Calculate the change in impedance spectrum under conditions with and without ultrasonic disturbance. From the electrochemical response signal and impedance spectrum change Extracting feature vectors ; S6 will use the feature vector Input a pre-trained classification-regression joint model and output the defect category and residual stress value.

8. The non-destructive testing method for micro-regions of materials based on physical field perturbation according to claim 7, characterized in that, After applying a constant potential in step S2, current drift must be satisfied. And continue Under the given conditions, initial electrochemical impedance spectroscopy was then acquired. When performing potentiodynamic scanning, the current sampling frequency is set to... .

9. The non-destructive testing method for micro-regions of materials based on physical field perturbation according to claim 7, characterized in that, In step S3, the ultrasonic dwell time of the ultrasonic excitation module is set to... It works in conjunction with a robotic arm to scan the workpiece being measured. The robotic arm's step distance is... Monitoring during the scanning process and The fluctuation range, when the fluctuation range is all less than The system determines the signal to be valid when the current exceeds the range. If the current exceeds the range, the system will interrupt the scan and prompt the user to check the coupling status.

10. The non-destructive testing method for micro-regions of materials based on physical field perturbation according to claim 7, characterized in that, In step S5 Calculated by fitting electrochemical impedance spectroscopy. Obtained analytically through equivalent circuit model. For microcurrent transients The rate of change; eigenvectors It also includes the transient peak change rate. Relaxation time signal rise time and the spectral energy of specific frequency bands Relaxation time The fitting equation is obtained by solving the transient waveform of the microcurrent using a double exponential decay fitting method. , Weighted average relaxation time Spectral energy of a specific frequency band By analyzing the electrochemical response signal The integral equation is obtained by performing a Fourier transform and integrating within the characteristic frequency band. ,in , , This represents the amplitude of the spectrum after Fourier transform.

Images