Crack detection system and method

Abstract

The application relates to the field of nondestructive testing of special equipment and robot technology, in particular to a crack detection system and method, which comprises a mobile bearing platform used for adsorbing and moving along the surface of a pipeline to be detected; a composite sensor probe rod connected to the mobile bearing platform and used for emitting ultrasonic signals to a region to be detected and receiving echo signals; a flexible coupling mechanism arranged at the bottom of the composite sensor probe rod and used for fitting the curved surface of the pipeline to be detected; and a control unit used for controlling the mobile bearing platform to move along a planned path according to feedback information of a positioning and mapping module, controlling the composite sensor probe rod to perform a detection operation, and controlling the flexible coupling mechanism to maintain a preset contact pressure during detection. Therefore, the problems that a three-way complex curved surface has many blind areas, a micro-crack detection rate is low, a manual dependence is high and it is difficult to adapt to nickel-based materials are solved, and full-automatic, high-sensitivity, blind-area-free accurate detection and evaluation are realized.

Description

Technical Field

[0001] This application relates to the field of nondestructive testing of special equipment and robotics, and in particular to a crack detection system and method. Background Technology

[0002] Due to their geometric discontinuities, tee fittings in power plant piping systems are prone to significant stress concentrations at the shoulder (R1 zone) and the root of branch pipe connections (R2 zone). During long-term service, thermal shock can easily induce microcracks, potentially leading to catastrophic accidents. As thermal power technology advances towards higher parameters, future 700℃-class ultra-supercritical units will utilize new materials such as nickel-based high-temperature alloys. Their anisotropy and complex acoustic attenuation characteristics place higher demands on detection technologies, making existing methods difficult to apply directly.

[0003] In related technologies, current routine non-destructive testing of in-service tees mainly relies on manual ultrasonic testing, which has significant drawbacks. Firstly, the abrupt curvature changes in tees cause severe scattering or ineffective incidentness of traditional ultrasonic beams. While phased array ultrasonic testing can be improved through electronic deflection, its fixed-angle wedges cannot perfectly match complex curved surfaces, leading to focus drift and incomplete coverage, resulting in a high rate of missed detections. Secondly, traditional piezoelectric ultrasonic probes suffer from poor coupling stability, significant signal attenuation, and insufficient resolution of sub-millimeter-level shallow microcracks. Furthermore, the testing process is cumbersome, heavily reliant on experienced personnel from grinding and couplant application to manual interpretation, with each test taking several hours. This fails to meet the practical demands of short maintenance windows and high efficiency requirements in power plants, necessitating a solution. Summary of the Invention

[0004] This application provides a crack detection system and method to solve the problems of numerous blind spots, low microcrack detection rate, high dependence on manual operation, and difficulty in adapting to nickel-based materials in the detection of complex curved surfaces of tees. It achieves fully automatic, highly sensitive, and blind-spot-free accurate detection and evaluation.

[0005] The first aspect of this application provides a crack detection system, including: A mobile carrier platform, which integrates a positioning and mapping module, is configured to adsorb and move along the surface of the pipe to be inspected. A composite sensor probe is connected to the mobile support platform and is configured to emit ultrasonic signals to the area to be detected and receive echo signals during movement. A flexible coupling mechanism is disposed at the bottom of the composite sensor probe, and the flexible coupling mechanism is configured to conform to the curved surface of the pipe to be tested to establish acoustic coupling; The control unit is connected to the mobile support platform, the composite sensor probe, and the flexible coupling mechanism respectively. The control unit is configured to control the mobile support platform to move along the planned path according to the feedback information of the positioning and mapping module, control the composite sensor probe to perform detection operations, and control the flexible coupling mechanism to maintain a preset contact pressure during detection.

[0006] Optionally, in some embodiments, the composite sensor probe includes: The guided wave primary screening module is configured to transmit guided wave signals to the area to be detected and receive reflected echoes to generate primary screening detection data; The focused detection module is configured to emit a programmable focused ultrasound signal toward the area to be detected and receive reflected echoes to obtain target detection data.

[0007] Optionally, in some embodiments, the waveguide screening module is composed of multiple piezoelectric wafers, and the focusing detection module includes a permanent magnet array, a telescopic substrate, and a circuit layer etched with tortuous coils.

[0008] Optionally, in some embodiments, the control unit has a built-in stress knowledge base and a focus compensation engine; The stress knowledge base stores a parametric tee model and the theoretical stress distribution data corresponding to the parametric tee model. The focusing compensation engine is configured to calculate the emission delay of each array element in the focusing detection module based on the local curvature information of the area to be inspected, using a preset delay algorithm, so as to form a focused sound beam in the curved area.

[0009] Optionally, in some embodiments, the launch delay calculation formula performed by the focusing compensation engine is: ; in, For the transmission delay of the i-th element, The position vector of the array element. Let the focus position vector be... For the transverse wave sound velocity of the material, The curvature compensation coefficient is related to the material and frequency. , Let be the two principal radii of curvature of the workpiece surface at the focal point.

[0010] Optionally, in some embodiments, the flexible coupling mechanism includes: The double-layer membrane structure includes an inner sealed air chamber and an outer contact layer; An air pump is configured to regulate the air pressure in the inner sealed air chamber; The outer contact layer is covered with a conductive mesh layer, which is grounded to discharge eddy currents generated during the detection process.

[0011] The crack detection system provided in this application embodiment uses a mobile support platform carrying a composite sensor probe. Under the unified scheduling of the control unit, the system automatically plans the detection path, utilizes a guided wave rapid screening module for efficient scanning and generates an attenuation-dispersion joint distribution spectrum, identifies suspicious areas, and then drives the focusing module to move above the suspicious areas. It dynamically calculates the emission delay of each array element to achieve curvature-adaptive precise focusing scanning. A multi-angle measurement method is used to accurately quantify the cracks and assess their remaining lifespan. This solves the problems of numerous blind spots, low microcrack detection rate, high reliance on manual intervention, and difficulty in adapting to nickel-based materials in detecting complex curved surfaces of tee tunnels, achieving fully automatic, high-sensitivity, and blind-spot-free precise detection and assessment.

[0012] A second aspect of this application provides a crack detection method applied to the crack detection system described above, wherein the method includes the following steps: The control unit generates a detection path based on the parameterized tee model and corresponding theoretical stress distribution data stored in the stress knowledge base, and controls the mobile bearing platform to move along the detection path. During the movement of the mobile carrier platform, the control unit controls the guided wave screening module to emit guided wave signals to the area to be detected, receives reflected echoes to generate preliminary screening data, and generates an attenuation-dispersion joint distribution spectrum based on the preliminary screening data. The characteristic parameters of the attenuation-dispersion joint distribution map are input into a preset recognition model to obtain the doubt confidence of the region to be detected. Regions with doubt confidence greater than a preset threshold are marked as doubtful regions, and the curvature information of the doubtful regions is extracted. The control unit controls the mobile carrier platform to move the focusing detection module above the suspicious area, and the focusing compensation engine calculates the emission delay of each array element based on the curvature information, and controls the focusing detection module to perform focusing scanning to obtain target detection data; The target detection data is processed to calculate the geometric parameters of the crack, and the remaining life is assessed based on the geometric parameters, and the detection results are output.

[0013] According to one embodiment of this application, generating an attenuation-dispersion joint distribution map based on the initial screening detection data includes: Based on the reflected echo received by the waveguide screening module, the characteristic parameters of each detection point are extracted; The characteristic parameters of each detection point are associated with the detection location information to generate the attenuation-dispersion joint distribution map. The characteristic parameters of each detection point include at least one of local attenuation, dispersion curve slope, and reflected wavelet energy.

[0014] According to one embodiment of this application, processing the target detection data, calculating the geometric parameters of the crack, and performing a remaining lifetime assessment based on the geometric parameters includes: The target detection data is processed to determine the crack endpoint, and the projected length of the crack is calculated based on the crack endpoint. The focused detection module emits focused ultrasonic signals toward the crack at multiple different incident angles, measures the reflection echo time difference at each angle, and obtains the true depth of the crack based on the reflection echo time difference at each incident angle. The remaining life loss of the pipeline under inspection is calculated based on the crack projection length and the actual depth.

[0015] According to one embodiment of this application, the actual depth of the crack is: ; in, This represents the true depth of the crack. For the longitudinal wave velocity of the material, Let be the time difference of the reflected echo at the j-th incident angle. The angle of incidence, The angle between the normals of the crack surface and the crack surface.

[0016] According to the crack detection method provided in this application, a composite sensor probe is mounted on a mobile support platform. Under the unified scheduling of the control unit, the detection path is automatically planned. A guided wave rapid screening module is used for efficient scanning and generates an attenuation-dispersion joint distribution spectrum to identify suspicious areas. Then, the focusing module is driven to move above the suspicious areas, and the emission delay of each array element is dynamically calculated to achieve curvature-adaptive precise focusing scanning. A multi-angle measurement method is used to accurately quantify the crack and assess the remaining life. This solves the problems of numerous blind spots, low microcrack detection rate, high reliance on manual intervention, and difficulty in adapting to nickel-based materials in detecting complex curved surfaces of tee tunnels. It achieves fully automatic, high-sensitivity, and blind-spot-free accurate detection and assessment.

[0017] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0018] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1This is a flowchart illustrating the overall structure and workflow of a crack detection system according to an embodiment of this application. Figure 2 A detailed logic flowchart of an adaptive focusing algorithm engine according to an embodiment of this application; Figure 3 This is a schematic flowchart of a crack detection method according to an embodiment of this application. Detailed Implementation

[0019] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0020] The crack detection system and method of this application are described below with reference to the accompanying drawings. Addressing the problems mentioned in the background art, such as numerous blind spots in the detection of tee areas, low microcrack detection rates, and high reliance on manual labor, this application provides a crack detection system and method that achieves near 100% blind-spot-free coverage detection of high-stress concentration areas such as R1 and R2 of tees. It ensures high sensitivity detection and accurate quantification of internal wall microcracks with a depth ≥0.8mm. By constructing a fully automated closed-loop detection process, the single detection time is controlled within 45 minutes, significantly reducing labor costs and skill dependence. Through modular parameter design, the system can be extended to the detection of tees made of new materials such as nickel-based high-temperature alloys, laying a technical foundation for the detection of nickel-based material pipelines in next-generation 700℃ units.

[0021] Specifically, the crack detection system includes a mobile support platform, a composite sensor probe, a flexible coupling mechanism, and a control unit.

[0022] The system comprises: a mobile support platform integrating a positioning and mapping module, configured to adhere to and move along the surface of the pipe to be inspected; a composite sensor probe connected to the mobile support platform, configured to emit ultrasonic signals to the area to be inspected and receive echo signals during movement; a flexible coupling mechanism located at the bottom of the composite sensor probe, configured to conform to the curved surface of the pipe to be inspected to establish acoustic coupling; and a control unit connected to the mobile support platform, the composite sensor probe, and the flexible coupling mechanism, configured to control the mobile support platform to move along a planned path based on feedback information from the positioning and mapping module, control the composite sensor probe to perform inspection operations, and control the flexible coupling mechanism to maintain a preset contact pressure during inspection.

[0023] Optionally, the mobile support platform in this embodiment can be a four-wheeled wall-climbing robot with a permanent magnet adsorption array, a single-wheel adsorption force design value ≥200N, and an integrated high-precision laser SLAM module and inertial measurement unit (IMU) to achieve stable crawling and real-time centimeter-level positioning on the pipe wall. This platform has the extended capability to operate on nickel-based (non-ferromagnetic) material surfaces through additional vacuum adsorption or auxiliary magnetic tracks.

[0024] Specifically, a mobile support platform is used to adhere to and move along the pipe surface, integrating positioning and mapping modules. A composite sensor probe, connected to the mobile support platform, integrates a guided wave screening module and a focusing detection module. A flexible adaptive coupling mechanism, encapsulated at the bottom of the composite sensor probe, ensures stable acoustic coupling between the probe and the curved pipe wall. The control unit has a built-in tee-specific knowledge base and a real-time focusing compensation engine. The tee-specific knowledge base pre-stores tee CAD models and their theoretical stress distribution data. The real-time focusing compensation engine dynamically calculates the emission delay of each element in the focusing detection module based on the local curvature information of the point to be inspected, using a curvature compensation algorithm to achieve precise focusing on complex curved areas.

[0025] Optionally, in some embodiments, the composite sensor probe includes: a guided wave screening module configured to emit guided wave signals toward the area to be detected and receive reflected echoes to generate preliminary screening detection data; and a focused detection module configured to emit programmable focused ultrasonic signals toward the area to be detected and receive reflected echoes to obtain target detection data.

[0026] Specifically, in this embodiment of the application, the composite sensor probe is rigidly connected to the front end of the robot chassis, and integrates the guided wave screening module and the focusing detection module in a parallel layout.

[0027] Optionally, in some embodiments, the waveguide screening module consists of multiple piezoelectric wafers, and the focusing detection module includes a permanent magnet array, a telescopic substrate, and a circuit layer etched with tortuous coils.

[0028] Furthermore, the guided wave screening module can be composed of a linear array of multiple (e.g., eight) independently packaged broadband piezoelectric ceramic wafers, automatically switching the excitation mode according to different detection areas. This guided wave screening module is programmable and can excite a T(0,1) mode torsional wave centered at 32kHz in the main pipe area and an L(0,2) mode longitudinal wave centered at 80kHz in the branch pipe area, based on the robot's position information, achieving optimal detection mode matching for different areas. For nickel-based materials, the excitation frequency and modal parameters can be preset according to the material's acoustic properties.

[0029] The focusing detection module, serving as the core precision inspection probe, employs a multi-layered composite structure. The bottom layer is an N52 neodymium iron boron permanent magnet array, providing a static bias magnetic field perpendicular to the tube wall. The middle layer is an oscillator substrate made of Terfenol-D magnetostrictive material. The surface layer is a 16-channel rectangular zigzag coil array fabricated using photolithography, with a programmable operating frequency ranging from 2 to 10 MHz. When detecting nickel-based materials, this module can switch to an electromagnetic ultrasonic resonance mode to overcome the effects of weak magnetism.

[0030] Optionally, in some embodiments, the flexible coupling mechanism includes: a double-layer membrane structure including an inner sealed air chamber and an outer contact layer; an air pump configured to regulate the air pressure of the inner sealed air chamber; wherein the surface of the outer contact layer is covered with a conductive mesh layer, the conductive mesh layer being grounded for discharging eddy currents generated during the detection process.

[0031] Specifically, the flexible coupling mechanism is encapsulated at the bottom of the sensor probe and is an inflatable double-layer silicone rubber membrane structure. The inner layer is a sealed air chamber, which is maintained at a constant pressure of 0.18±0.02MPa by a micro air pump to ensure uniform and stable acoustic coupling between the probe and the curved tube wall. The outer layer is coated with a high-density (120 mesh) diamond-shaped copper mesh by etching and is reliably grounded. Its surface resistivity is less than 0.1Ω, which is used to discharge the eddy currents generated on the tube wall when the EMAT (Effective Detection Module) is working, significantly improving the signal-to-noise ratio.

[0032] Optionally, in some embodiments, the control unit has a built-in stress knowledge base and a focusing compensation engine; wherein, the stress knowledge base stores a parameterized tee model and theoretical stress distribution data corresponding to the parameterized tee model; the focusing compensation engine is configured to calculate the emission delay of each array element in the focusing detection module according to the local curvature information of the area to be inspected through a preset delay algorithm, so as to form a focused sound beam in the curved area.

[0033] Specifically, the control unit in this application embodiment is the control core of the system, and it has a built-in three-way exclusive knowledge base and a real-time focus compensation engine.

[0034] The dedicated tee knowledge base pre-stores parametric CAD models, including equal-diameter tees and reducing tees. Each model not only includes precise geometric dimensions (D, d, R1, R2, etc.) but also embeds typical stress distribution cloud maps obtained from mature pressure area methods or finite element analysis, thus pre-marking theoretically high-risk areas. For future 700℃ units, high-temperature creep fatigue damage models of nickel-based materials can be imported to predict potential crack initiation zones. The core of the real-time focusing compensation engine is the introduction of a curvature-aware dynamic delay compensation algorithm.

[0035] During EMAT phased array focusing, the transmission delay of each array element is... Determined by the standard sound path delay and an original compensation term, its mathematical model is as follows: ; in, and Let be the position vector of the i-th element and the focus position vector, respectively. The transverse wave velocity in the material. Curvature compensation term. The key to overcoming the geometric blind spot in this invention is the calculation formula: ; In this formula, , These are the two principal radii of curvature of the workpiece surface at the focal point, extracted in real time from the currently matched CAD model. It is a material-frequency related curvature compensation coefficient, obtained through extensive experimental calibration. For commonly used steels in power plants, such as P91 steel at a frequency of 5MHz, ≈ 0.017; P92 steel under the same conditions, ≈ 0.019; for the nickel-based alloy Inconel 740H, after calibration ≈ 0.022 (5MHz). This compensation term corrects for the additional phase delay that occurs when the acoustic beam wavefront propagates on a non-planar surface (especially a concave surface), thus ensuring accurate focusing in regions of abrupt curvature changes.

[0036] Therefore, the formula for calculating the launch delay performed by the focus compensation engine is: ; in, For the transmission delay of the i-th element, The position vector of the array element. Let the focus position vector be... For the transverse wave sound velocity of the material, The curvature compensation coefficient is related to the material and frequency. , Let be the two principal radii of curvature of the workpiece surface at the focal point.

[0037] Therefore, the crack detection system of this application is particularly suitable for the detection, quantification, and safety assessment of microcrack clusters initiating in high stress concentration areas caused by abrupt geometric changes on the inner surfaces of tees in pressure-bearing components such as headers, main steam pipelines, and feedwater pipelines in thermal power plants. This invention can also be extended to the in-service inspection of nickel-based high-temperature alloy pipeline tees in next-generation 700℃-class ultra-supercritical units, providing technical reserves for the safe operation of units with higher parameters.

[0038] To facilitate a clearer and more intuitive understanding of the crack detection system of this application by those skilled in the art, the following is combined with... Figure 2 and Figure 3 Please provide a detailed explanation.

[0039] like Figure 2 As shown, the physical module consists of a mobile support platform, a composite sensor probe, a flexible adaptive coupling mechanism, and a collaborative control and processing unit, forming the hardware foundation of the system. The detection process comprises four steps: guided wave initial screening, suspicious area identification, EMAT adaptive focusing scanning, and crack quantification and life assessment, forming a complete automated detection closed loop.

[0040] Furthermore, the detailed logic flow of the adaptive focusing algorithm engine is as follows: Figure 3 As shown, the system first inputs the focal coordinates and material type, extracts the two principal curvature radii at the focal point from the CAD model, and queries the corresponding curvature compensation coefficients based on the material type, then calculates the curvature compensation term. The system sequentially iterates through the 16 array elements, calculating the standard path delay between each element and the focal point, and adding this delay to the curvature compensation term to obtain the total transmission delay for that element. Finally, a complete 16-channel delay sequence is output. By introducing the curvature compensation term, dynamic correction of the beam focusing delay in the complex curved surface region of the three-way junction is achieved.

[0041] The crack detection system proposed in this application uses a mobile support platform carrying a composite sensor probe. Under the unified scheduling of the control unit, the system automatically plans the detection path, utilizes a guided wave rapid screening module for efficient scanning and generates an attenuation-dispersion joint distribution spectrum, identifies suspicious areas, and then drives the focusing module to move above the suspicious areas. It dynamically calculates the emission delay of each array element to achieve curvature-adaptive precise focusing scanning. A multi-angle measurement method is used to accurately quantify the cracks and assess their remaining lifespan. This solves the problems of numerous blind spots, low microcrack detection rate, high reliance on manual intervention, and difficulty in adapting to nickel-based materials in detecting complex curved surfaces of tee tunnels, achieving fully automatic, high-sensitivity, and blind-spot-free accurate detection and assessment.

[0042] Next, referring to the accompanying drawings, a crack detection method according to an embodiment of this application is described, which is applied to the crack detection system described above.

[0043] like Figure 3 As shown, the crack detection method includes the following steps: In step S301, the control unit generates a detection path based on the parameterized tee model and corresponding theoretical stress distribution data stored in the stress knowledge base, and controls the mobile bearing platform to move along the detection path.

[0044] In step S302, during the movement of the mobile carrier platform, the control unit controls the guided wave primary screening module to emit guided wave signals to the area to be detected, receives the reflected echoes to generate primary screening detection data, and generates an attenuation-dispersion joint distribution spectrum based on the primary screening detection data.

[0045] Specifically, adaptive path planning and guided wave dual-mode initial screening are performed. The control unit loads the CAD model of the tee to be inspected and automatically generates the optimal detection path (e.g., main pipe helix + branch pipe axial scan) based on its geometric features and embedded stress distribution map. The robot moves along the path and uses the guided wave initial screening module to excite dual-mode guided waves according to a preset protocol, while simultaneously acquiring full waveform data. The system calculates and plots the attenuation-dispersion joint distribution spectrum of the entire region. For nickel-based materials, the guided wave excitation parameters automatically call the corresponding dispersion curve data from the knowledge base to ensure modal purity.

[0046] Furthermore, in some embodiments, generating an attenuation-dispersion joint distribution map based on the initial screening detection data includes: extracting characteristic parameters of each detection point based on the reflected echo received by the guided wave initial screening module; associating the characteristic parameters of each detection point with the detection location information to generate an attenuation-dispersion joint distribution map; wherein the characteristic parameters of each detection point include at least one of local attenuation, dispersion curve slope, and reflected wavelet energy.

[0047] In step S303, the characteristic parameters of the attenuation-dispersion joint distribution map are input into the preset identification model to obtain the doubt confidence of the area to be detected. Areas with doubt confidence greater than the preset threshold are marked as doubtful areas, and the curvature information of the doubtful areas is extracted.

[0048] The deep learning model is a convolutional neural network trained on historical detection data, and its input features include local attenuation, dispersion curve slope, and specific reflected wavelet energy.

[0049] Specifically, machine learning-based intelligent identification of suspicious areas is performed. The feature parameters of the obtained distribution map (such as local attenuation Att, dispersion curve slope Sd, and specific reflected wavelet energy Er) are input into a pre-trained deep learning classification model (e.g., a convolutional neural network trained on a large amount of historical tee detection data, which may include transfer learning models of nickel-based material samples). This model outputs a confidence score Ps related to the probability of stress concentration / crack presence. When Ps exceeds a preset threshold (e.g., 92%), the system marks the area as a first-level suspicious area and binds its spatial coordinates to the curvature information (R1, R2) in the CAD model.

[0050] In step S304, the control unit controls the mobile carrier platform to move the focusing detection module above the suspicious area, and the focusing compensation engine calculates the emission delay of each array element based on the curvature information, and controls the focusing detection module to perform focusing scanning to obtain target detection data.

[0051] Specifically, CAD model-guided EMAT adaptive focusing scans are performed. For each marked first-level suspicious region, the control unit drives the focusing detection module to move above the suspicious region. The focusing engine, based on the region's coordinates Pf and the bound curvature (R1, R2), combined with material coefficients... The precise transmission delay sequence of 16 channels is calculated in real time using the aforementioned delay calculation formula, forming a focused acoustic beam that can adaptively conform to the local curvature of the point. Subsequently, a high-density grid scan is performed on the area in steps ≤0.5mm to obtain a high signal-to-noise ratio echo signal (i.e., target detection data), where the acoustic beam electronic deflection range can reach ±85°, ensuring complete coverage. When detecting nickel-based materials, the EMAT operating frequency can be adjusted to the optimal signal-to-noise ratio range (e.g., 3-7MHz), and pulse compression technology can be enabled to improve resolution.

[0052] In step S305, the target detection data is processed, the geometric parameters of the crack are calculated, the remaining life is assessed based on the geometric parameters, and the detection results are output.

[0053] Furthermore, in some embodiments, the target detection data is processed to calculate the geometric parameters of the crack and to assess the remaining life based on the geometric parameters. This includes: processing the target detection data to determine the crack endpoints and calculating the projected length of the crack based on the crack endpoints; using a focused detection module to emit focused ultrasonic signals to the crack at multiple different incident angles, measuring the reflection echo time difference at each angle, and obtaining the true depth of the crack based on the reflection echo time difference at each incident angle; and calculating the remaining life loss of the pipeline under inspection based on the crack projected length and the true depth.

[0054] Specifically, precise quantitative analysis of cracks and assessment of remaining life are performed. The acquired high signal-to-noise ratio scanning signal clusters are processed, the echo signals are analyzed, and a multi-angle measurement method is used to calculate the true crack depth. Combined with the crack length, the remaining life is automatically assessed, and a test report is generated.

[0055] Furthermore, the crack endpoints were determined using the 6dB drop method, thereby calculating the projected length Lcrack. This was achieved by varying the incident angle of the focused acoustic beam (in at least three different directions) and measuring the time difference of the reflected echoes. The true depth d of the crack is calculated using the following formula: Basic measurement formula (single incident angle αj): ; in, The longitudinal wave velocity in the material, θ The angle between the normal direction of the crack surface and the normal of the surface at the detection point (unknown quantity).

[0056] Furthermore, the solution formula for calculating the true crack depth using the multi-angle measurement method (simultaneous multi-angle): ; in, The true depth of the crack. For the longitudinal wave velocity of the material, Let be the time difference of the reflected echo at the j-th incident angle. The angle of incidence, The angle between the normals of the crack surface, .

[0057] In the actual iterative calculation, an optimization algorithm is used to find a set of (d, θ) that minimizes the overall error between the calculated value and the measured value Δtj when substituted into the equations for each angle. The final output d is the true depth of the crack.

[0058] Finally, the quantified crack size (Lcrack, d) is substituted into the fatigue crack propagation model to automatically calculate the percentage of remaining life loss or safety factor caused by the crack, and generate a digital inspection report containing all details. For nickel-based materials, a high-temperature creep-fatigue crack propagation model based on that material can be built-in to achieve life assessment.

[0059] Furthermore, the crack detection method of this application includes extended applications for nickel-based superalloy materials: the tee's dedicated knowledge base pre-stores acoustic parameters, curvature compensation coefficients, and high-temperature creep fatigue damage models of nickel-based materials, and automatically calls the corresponding parameters during the detection process to realize the detection and evaluation of nickel-based alloy pipeline tees for next-generation 700℃ ultra-supercritical units.

[0060] To help those skilled in the art to better understand the crack detection system and method of this application, a detailed description is provided below with examples.

[0061] As an embodiment of this application, an in-service inspection of a P92 steel reducing tee (DN550×DN300) for the main steam pipeline of a 600MW supercritical thermal power unit was conducted.

[0062] (1) Preliminary preparation and deployment: Remove the external insulation layer of the tee. Use light sandblasting to treat the surface of the test area to a roughness Ra≈6.3μm, without needing to achieve a mirror finish. Attach optical positioning targets to the main and branch pipes of the tee at 500mm intervals. Place the robot system on the main pipeline, start self-inspection and SLAM mapping, and the system automatically identifies and matches the pre-stored corresponding CAD model (including stress distribution cloud map).

[0063] (2) Initial screening: The robot automatically moves along the planned path (main pipe spiral + branch pipe axial scan). The waveguide module excites a 32kHz T(0,1) wave in the main pipe section and switches to an 80kHz L(0,2) wave in the branch pipe section. The system generates an attenuation heat map in real time. The built-in AI classifier (based on an improved ResNet50) analyzes the map and identifies an elliptical region at the root of the branch pipe (coordinates X=1200mm, Y=650mm), with an average attenuation value Att=18dB, a dispersion slope change of 25%, and a confidence level Ps=96%, which is marked as a first-level suspicious area.

[0064] (3) Perform fine inspection: The EMAT probe moves to above the suspicious area. The control unit reads the curvature radius of this point R1=40mm, R2=25mm from the CAD model, and selects the P92 steel compensation coefficient kc=0.019. The focusing engine calculates the delay parameters and drives the probe to perform a 0.5mm step focusing scan of this area, with the sound beam deflected ±60° to cover the entire elliptical area. During the scanning process, the flexible coupling mechanism automatically adjusts the air pressure to ensure good contact between the probe and the curved surface.

[0065] (4) Results Analysis and Report: Clear crack reflection signals were obtained from the scan, with a signal-to-noise ratio of 42dB. The crack projection length was measured to be 6.8mm using the 6dB drop method. By fitting scan data from three different angles (0°, 20°, -20°), the crack depth d=2.1mm was calculated, with an angle θ=12° between the crack and the crack surface normal. The system automatically invoked the fatigue crack propagation model for P92 steel and calculated that the crack had caused approximately 12.7% of the lifespan loss. The final report presents the crack location, size, C-scan image, and evaluation conclusions in a three-dimensional visualization format.

[0066] As another embodiment of this application, a simulation test is conducted on nickel-based alloy pipeline tees for 700°C ultra-supercritical units.

[0067] To verify the adaptability of this invention to the next generation of high-parameter units, a simulated tee test block of Inconel740H material (main pipe outer diameter 600mm, wall thickness 80mm, branch pipe outer diameter 350mm) was prepared under laboratory conditions, and microcracks with a depth of 1.0mm and a length of 5mm were artificially processed in its R1 region.

[0068] (1) Knowledge base configuration: The material acoustic parameters of Inconel 740H (longitudinal wave velocity 5820m / s, transverse wave velocity 3100m / s, curvature compensation coefficient kc=0.022 at 5MHz) and typical stress distribution (obtained through high temperature creep finite element analysis) are entered into the T-junction exclusive knowledge base.

[0069] (2) Detection execution: The robot system scans along a preset path. The waveguide module adjusts the excitation frequency to 25kHz (T(0,1)) and 60kHz (L(0,2)) to take into account the high acoustic attenuation characteristics of nickel-based materials, and uses pulse superposition technology to improve the signal-to-noise ratio. The AI ​​model (through transfer learning from a small number of nickel-based material samples) successfully identified suspicious points in the R1 region (confidence level 94%).

[0070] (3) EMAT fine inspection: The focusing engine calculates the delay based on the radius of curvature (R1=35mm, R2=22mm) and compensation coefficient, driving the EMAT probe to perform focusing scanning at a center frequency of 5MHz. To overcome the weak magnetism of nickel-based materials, the EMAT working mode is switched to the Lorentz force mechanism, and pulse compression technology is enabled, ultimately achieving a crack reflection signal-to-noise ratio of 38dB. The crack depth was calculated to be 1.02mm through multi-angle measurements, which is in good agreement with the preset value.

[0071] (4) Life assessment: The system calls the built-in nickel-based material creep-fatigue crack propagation model and, combined with the design and operating parameters of the tee (temperature 700℃, pressure 35MPa), assesses that the crack is expected to operate safely for about 32,000 hours under the current operating conditions, providing a scientific basis for maintenance decisions.

[0072] This embodiment demonstrates that, through modular parameter design, the present invention can be seamlessly extended to the testing of nickel-based material pipe tees for next-generation high-parameter units, possessing good forward-looking and technological reserve value.

[0073] In summary, the present invention can achieve at least the following beneficial effects: (1) It pioneers a new paradigm of model-guided targeted inspection: Unlike the "indiscriminate" full-domain scanning or simple anisotropic comparison of pipelines in correlation, this invention is the first to deeply integrate the theoretical stress distribution model of the tee with real-time detection data, so that the inspection behavior has clear targeting from the beginning, which greatly improves the inspection efficiency and the probability of defect detection. At the same time, the knowledge base can be extended to new materials such as nickel-based materials, laying the groundwork for future 700℃ unit inspection.

[0074] (2) Overcoming the technical bottleneck of adaptive focusing on complex curved surfaces: The original focusing delay algorithm with curvature compensation term δcorrection is the core solution to the acoustic blind zone in the R1 and R2 areas of the three-way junction. This algorithm enables the sound beam to dynamically adapt to local geometric changes, achieving a qualitative leap from "undetectable" or "barely detectable" to "precise focused detection". Experiments show that this algorithm can increase the focus energy by more than 6dB on a concave surface with a curvature radius of 40mm.

[0075] (3) A complete automated chain of “screening-location-quantification-assessment” has been constructed: This invention seamlessly integrates guided wave initial screening, AI recognition, EMAT fine inspection, automatic quantification and life assessment into a single robotic platform, forming an end-to-end solution. This not only significantly improves efficiency (single test time is reduced by about 80%), but also improves the consistency and reliability of test results by reducing human intervention.

[0076] (4) Provides scalability for future high-parameter units: The present invention fully considers the diversity of materials in the system design. Through the material parameter library and configurable detection process, it can be directly applied to the detection of nickel-based alloy pipe tees in the next generation of 700℃ ultra-supercritical units. In view of the characteristics of weak magnetism and high acoustic attenuation of nickel-based materials, the system can maintain excellent detection performance by adjusting the EMAT working mode, optimizing the guided wave frequency and compensation coefficient, etc., providing technical guarantee for the safe operation of higher parameter units.

[0077] (5) Verifiable superior performance indicators: On-site testing of the prototype system based on this scheme at the power plant showed that its detection coverage of the stress concentration zone of the three-way valve was increased from less than 50% by traditional methods to over 98%; the signal-to-noise ratio for detecting 0.8mm deep microcracks was improved by more than 8dB; and the single-point detection time was reduced by about 80%. Preliminary research tests on nickel-based material specimens showed that the signal-to-noise ratio for simulated cracks with a depth of 1.0mm could reach over 35dB, verifying its expansion potential.

[0078] It should be noted that the foregoing explanation of the crack detection system embodiment also applies to the crack detection method of this embodiment, and will not be repeated here.

[0079] According to the crack detection method proposed in this application, a composite sensor probe is mounted on a mobile support platform. Under the unified scheduling of the control unit, the detection path is automatically planned. A guided wave rapid screening module is used for efficient scanning and generates an attenuation-dispersion joint distribution spectrum to identify suspicious areas. Then, the focusing module is driven to move above the suspicious areas, and the emission delay of each array element is dynamically calculated to achieve curvature-adaptive precise focusing scanning. A multi-angle measurement method is used to accurately quantify the crack and assess the remaining life. This solves the problems of numerous blind spots, low microcrack detection rate, high reliance on manual intervention, and difficulty in adapting to nickel-based materials in detecting complex curved surfaces of tee tunnels. It achieves fully automatic, high-sensitivity, and blind-spot-free accurate detection and assessment.

[0080] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0081] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0082] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0083] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0084] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0085] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0086] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0087] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A crack detection system, characterized in that, include: A mobile carrier platform, which integrates a positioning and mapping module, is configured to adsorb and move along the surface of the pipe to be inspected. A composite sensor probe is connected to the mobile support platform and is configured to emit ultrasonic signals to the area to be detected and receive echo signals during movement. A flexible coupling mechanism is disposed at the bottom of the composite sensor probe, and the flexible coupling mechanism is configured to conform to the curved surface of the pipe to be tested to establish acoustic coupling; The control unit is connected to the mobile support platform, the composite sensor probe, and the flexible coupling mechanism respectively. The control unit is configured to control the mobile support platform to move along the planned path according to the feedback information of the positioning and mapping module, control the composite sensor probe to perform detection operations, and control the flexible coupling mechanism to maintain a preset contact pressure during detection.

2. The system according to claim 1, characterized in that, The composite sensor probe includes: The guided wave primary screening module is configured to transmit guided wave signals to the area to be detected and receive reflected echoes to generate primary screening detection data; The focused detection module is configured to emit a programmable focused ultrasound signal toward the area to be detected and receive reflected echoes to obtain target detection data.

3. The system according to claim 2, characterized in that, The waveguide screening module consists of multiple piezoelectric wafers, and the focusing detection module includes a permanent magnet array, a telescopic substrate, and a circuit layer etched with tortuous coils.

4. The system according to claim 2, characterized in that, The control unit has a built-in stress knowledge base and a focus compensation engine; The stress knowledge base stores a parametric tee model and the theoretical stress distribution data corresponding to the parametric tee model. The focusing compensation engine is configured to calculate the emission delay of each array element in the focusing detection module based on the local curvature information of the area to be inspected, using a preset delay algorithm, so as to form a focused sound beam in the curved area.

5. The system according to claim 4, characterized in that, The launch delay calculation formula executed by the focusing compensation engine is as follows: ； in, For the transmission delay of the i-th element, The position vector of the array element. Let the focus position vector be... For the transverse wave sound velocity of the material, The curvature compensation coefficient is related to the material and frequency. , Let be the two principal radii of curvature of the workpiece surface at the focal point.

6. The system according to claim 1, characterized in that, The flexible coupling mechanism includes: The double-layer membrane structure includes an inner sealed air chamber and an outer contact layer; An air pump is configured to regulate the air pressure in the inner sealed air chamber; The outer contact layer is covered with a conductive mesh layer, which is grounded to discharge eddy currents generated during the detection process.

7. A crack detection method, characterized in that, Applied to the crack detection system as described in any one of claims 1-6, wherein the method comprises the following steps: The control unit generates a detection path based on the parameterized tee model and corresponding theoretical stress distribution data stored in the stress knowledge base, and controls the mobile bearing platform to move along the detection path. During the movement of the mobile carrier platform, the control unit controls the guided wave screening module to emit guided wave signals to the area to be detected, receives reflected echoes to generate preliminary screening data, and generates an attenuation-dispersion joint distribution spectrum based on the preliminary screening data. The characteristic parameters of the attenuation-dispersion joint distribution map are input into a preset recognition model to obtain the doubt confidence of the region to be detected. Regions with doubt confidence greater than a preset threshold are marked as doubtful regions, and the curvature information of the doubtful regions is extracted. The control unit controls the mobile carrier platform to move the focusing detection module above the suspicious area, and the focusing compensation engine calculates the emission delay of each array element based on the curvature information, and controls the focusing detection module to perform focusing scanning to obtain target detection data; The target detection data is processed to calculate the geometric parameters of the crack, and the remaining life is assessed based on the geometric parameters, and the detection results are output.

8. The method according to claim 7, characterized in that, The step of generating an attenuation-dispersion joint distribution map based on the initial screening data includes: Based on the reflected echo received by the waveguide screening module, the characteristic parameters of each detection point are extracted; The characteristic parameters of each detection point are associated with the detection location information to generate the attenuation-dispersion joint distribution map. The characteristic parameters of each detection point include at least one of local attenuation, dispersion curve slope, and reflected wavelet energy.

9. The method according to claim 7, characterized in that, The process of processing the target detection data, calculating the geometric parameters of the crack, and assessing the remaining life based on the geometric parameters includes: The target detection data is processed to determine the crack endpoint, and the projected length of the crack is calculated based on the crack endpoint. The focused detection module emits focused ultrasonic signals toward the crack at multiple different incident angles, measures the reflection echo time difference at each angle, and obtains the true depth of the crack based on the reflection echo time difference at each incident angle. The remaining life loss of the pipeline under inspection is calculated based on the crack projection length and the actual depth.

10. The method according to claim 9, characterized in that, The actual depth of the crack is: ； in, This represents the true depth of the crack. For the longitudinal wave velocity of the material, Let j be the time difference of the reflected echo at the j-th incident angle. The angle of incidence, The angle between the normals of the crack surface and the crack surface.

Images