A wind power flange multi-channel ultrasonic nondestructive intelligent detection method

Abstract

The present application relates to the technical field of flange ultrasonic testing, and particularly relates to a wind power flange multi-channel ultrasonic nondestructive intelligent testing method. According to the geometric size data and material grade of the flange to be tested, the incident angle, the focusing depth and the element delay sequence are determined, and multi-channel scanning and ultrasonic signal collection are performed accordingly; the echo amplitude is extracted and compared with the reference amplitude to obtain the amplitude deviation; the abnormal region set is dynamically identified and output based on the amplitude deviation and the geometric size data; the defect type is judged according to the center position, the spatial extension length, the maximum amplitude deviation intensity of the abnormal region and the geometric size data; and periodic statistical analysis is performed based on historical data, and the incident angle and the focusing depth are dynamically corrected, so as to reduce the dependence on fixed standards or the experience of operators, and to adapt to material batch differences and process fluctuations.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic testing technology for flanges, and specifically to a multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges. Background Technology

[0002] Wind turbine flanges are critical connecting components in wind turbine generator sets, and internal forging defects directly affect the overall structural safety and service life. Therefore, it is essential to conduct rapid and accurate non-destructive testing on wind turbine flanges before they leave the factory or before they enter service.

[0003] Currently, ultrasonic testing is one of the main methods for detecting internal defects in wind turbine flanges. Existing technology includes a method and system for detecting defects in flange forging, as disclosed in Chinese invention patent CN120559087A. This method establishes a high-precision layered sound path data model, generates targeted array element excitation control files, and employs a dual-array phased array system combined with a two-stage AI model for defect identification and classification, aiming to improve the detection rate and classification accuracy of minute defects.

[0004] However, after in-depth analysis, the existing technology still has the following shortcomings: First, the process of creating personalized models for each workpiece is cumbersome, resulting in low efficiency of inspection deployment and difficulty in matching the rapid production and inspection rhythm of wind power flanges in batches and assembly lines; Second, the inspection parameters generated are fixed and lack self-optimization capabilities, making it unable to adapt to the differences in acoustic characteristics caused by material batches and process fluctuations; Third, although AI models are introduced for classification, a closed loop of inspection-feedback-optimization is not constructed, making it impossible to continuously correct core inspection parameters using historical data, and the reliance on human experience or fixed standards remains high. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and solve the problems of low deployment efficiency, fixed parameters lacking self-optimization ability, and difficulty in continuous optimization in the context of rapid production and testing in batch and assembly line environments for wind power flange ultrasonic testing.

[0006] The technical solution adopted by this invention to solve its technical problem is: a multi-channel ultrasonic non-destructive intelligent inspection method for wind power flanges, comprising the following steps: S1, according to the geometric dimensions and material grade of the flange to be inspected, the flange to be inspected is divided into a main inspection area, a first-class secondary inspection area, and a second-class secondary inspection area; based on the transverse wave velocity determined by the material grade, and according to the acoustic refraction law, the incident angles of the main inspection area, the first-class secondary inspection area, and the second-class secondary inspection area are calculated respectively; the focal points of the main inspection area, the first-class secondary inspection area, and the second-class secondary inspection area are determined, and the sound path distance from the probe detection surface to each focal point along the corresponding ultrasonic main beam direction is calculated as the focusing depth.

[0007] S2. Based on each incident angle and the corresponding focusing depth, combined with the transverse wave velocity and the spatial coordinates of each array element, calculate the propagation time of the ultrasonic waves emitted by each array element to reach the corresponding focal point synchronously. Using the array element with the longest propagation time as the benchmark, calculate the emission delay value of the remaining array elements to form an array element delay sequence. Perform multi-channel scanning according to the determined incident angle, focusing depth and array element delay sequence to obtain an ultrasonic signal dataset.

[0008] S3. Extract the echo amplitude of each scanning point in the ultrasound signal dataset and compare it with the reference amplitude to obtain the amplitude deviation.

[0009] S4. Dynamically identify and output a set of abnormal regions based on amplitude deviation and geometric dimension data; the set of abnormal regions includes at least the center position, spatial extension length and maximum amplitude deviation intensity of each abnormal region.

[0010] S5. Determine the defect type based on the center location, spatial extension length, maximum deviation intensity, and geometric dimensions of each abnormal area.

[0011] S6. Associate and store the geometric dimensions, incident angle, focusing depth, defect type and abnormal area set of this inspection, and periodically optimize and correct the incident angle and focusing depth based on the statistical analysis of historical associated and stored data.

[0012] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention directly determines the ultrasonic testing parameters based on the geometric dimension data and material grade of the flange to be inspected, without the need for independent three-dimensional scanning, acoustic characteristic measurement and complex sound path modeling of each flange to be inspected, which significantly shortens the test preparation time and meets the rapid and batch testing needs of the wind power flange industrial production line.

[0013] 2. This invention associates and stores the geometric dimensions, incident angle, focusing depth, defect type, center position of abnormal area and maximum deviation intensity of each inspection, and performs periodic statistical analysis based on historical data to dynamically correct the inspection parameters. This reduces the reliance on fixed standards or operator experience and adapts to material batch differences and process fluctuations.

[0014] 3. This invention dynamically identifies abnormal regions based on amplitude deviation and geometric dimension data, and directly outputs a set of abnormal regions including the center position, spatial extension length, and maximum amplitude deviation intensity of each abnormal region. This provides a clear and reliable basis for subsequent defect type judgment and parameter optimization. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the detection method of the present invention.

[0017] Figure 2 This is a schematic diagram illustrating the process of determining the incident angle in this invention.

[0018] Figure 3 This is a schematic diagram of the process for determining the focusing depth in this invention.

[0019] Figure 4 This is a schematic diagram illustrating the process of determining the array element delay sequence in this invention.

[0020] Figure 5 This is a schematic diagram of the process for dynamically identifying and outputting a set of abnormal regions according to the present invention. Detailed Implementation

[0021] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention. Furthermore, it should be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale.

[0022] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the invention or its application or use. Techniques, methods, and apparatus known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and apparatus should be considered part of the specification.

[0023] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0025] The following description, in conjunction with the accompanying drawings, details the specific scheme of the multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges provided by this invention.

[0026] Please see Figure 1The present invention provides a flowchart of a multi-channel ultrasonic non-destructive intelligent testing method for wind power flanges, which includes the following steps: Step S1: Determine the incident angle, focusing depth and array element delay sequence of ultrasonic testing based on the geometric dimensions and material grade of the flange to be tested.

[0027] In this invention, the geometric dimension data includes the thickness of the flange to be inspected, the center coordinates of the bolt holes, and the diameter.

[0028] Since the thickness direction is the main direction for bearing axial loads and also the direction in which defects are prone to propagate, the thickness is used to divide the main inspection area in order to achieve uniform interlayer coverage.

[0029] Furthermore, since the edge of the bolt hole is a stress concentration area, it is prone to cracks. Therefore, a secondary inspection area is defined by the center position coordinates and diameter of the bolt hole to specifically detect potential defects around the hole.

[0030] Geometric dimensions can be obtained by reading the CAD design drawings, 3D model files, or by using a 3D scanner for actual measurement of the flange to be inspected.

[0031] The material grade is used to determine the transverse wave velocity. The transverse wave velocity is a fundamental physical parameter for subsequent calculations of sound path, angle of incidence, and focusing delay. The transverse wave velocity can be determined by consulting the corresponding national standard, industry standard, or technical data sheet provided by the material supplier.

[0032] For example, the typical transverse wave velocity values ​​for common wind turbine flange steels Q345E or 42CrMo are approximately 3230 m / s and 3250 m / s, respectively.

[0033] Once the geometric dimensions and transverse wave velocity are obtained, the incident angle, focusing depth, and array element delay sequence can be determined sequentially.

[0034] Please see Figure 2 Step S10, the process of determining the incident angle, is as follows: First, taking the thickness of each layer as... The total thickness of the flange to be inspected is divided into several equal layers. For each thickness layer, the direction of the main ultrasonic beam is set to extend along the thickness direction, forming the main inspection area.

[0035] in, Typically, the probe chip size is 1 to 2 times the actual size to ensure sufficient overlap of the main ultrasonic beam between adjacent focusing layers and to avoid missed detections. Meanwhile, It is less than one-third the near-field length of the phased array probe to ensure that the focusing effect is located in the far-field region, reducing the impact of the diffusion of the main ultrasonic beam on the detection sensitivity.

[0036] In addition to the main inspection area, two types of secondary inspection areas need to be defined for the bolt hole structure of the flange to be inspected.

[0037] Firstly, considering that the edge of the bolt hole is a region of concentrated processing and service stress, it is prone to defects such as radial cracks radiating outward from the hole wall or circumferential cracks distributed along the hole circumference. Therefore, an annular area formed by extending radially outward from the center of the bolt hole is designated as a secondary inspection area.

[0038] The set radial distance is typically 0.5 to 1.5 times the bolt hole diameter. In this invention, the bolt hole diameter can be used as the set radial distance, for example.

[0039] For bolt holes with larger diameters, considering that the influence range of edge stress may be wider, the set radial distance can be appropriately increased to 1.3 times the bolt hole diameter; for flanges with high preload, the set radial distance can also be increased to ensure coverage of potential microcrack propagation zones.

[0040] For each Class I secondary detection area, the direction of the main ultrasonic beam extends radially along the flange to be inspected and points to the center axis of the bolt hole at the corresponding scanning point.

[0041] Secondly, considering that the area between adjacent bolt holes is subjected to shear and tensile stress when the flange under inspection is subjected to force, it is prone to extended cracks that extend along the line connecting the holes, or defects that are distributed in a band, such as banded inclusions.

[0042] Therefore, a strip-shaped area of ​​a predetermined width extending to both sides, based on the line connecting the centers of adjacent bolt holes, is designated as the secondary detection area. For each secondary detection area, the direction of the main ultrasonic beam is perpendicular to the line connecting the centers of adjacent bolt holes.

[0043] Since the stress-affected area between holes is typically related to the hole diameter, using the bolt hole diameter as the single-sided extension width can reasonably cover the potential defect zone between adjacent holes. Therefore, the width is set to be twice the single-sided extension width, which is the total width of the strip-shaped area.

[0044] If the flange under inspection is operating in an environment with severe vibration or uneven loading, consider increasing the width of the single-sided extension to 1.5 times the bolt hole diameter to cover a wider range of potential defects.

[0045] After the region division is completed, based on the transverse wave sound velocity and the acoustic refraction law, the incident angles of the main detection region, the first-class secondary detection region, and the second-class secondary detection region are calculated respectively to form a set of incident angles.

[0046] The specific formula for calculating the angle of incidence is as follows: .

[0047] in, The incident angle of the probe in the coupling agent; The transverse wave velocity in the flange to be inspected; The required angle of the shear wave in the flange to be inspected.

[0048] The longitudinal wave velocity is the velocity of sound in the coupling agent; in this invention, the coupling agent is usually water or oil, with water corresponding to... Approximately 1480 m / s, corresponding to oils It is approximately 1400 m / s.

[0049] For the main detection area For Class I and Class II secondary detection areas, Calculate using the following formula: .

[0050] in, This is the direction vector of the main ultrasonic beam. To detect the surface normal vector.

[0051] Please see Figure 3 Step S11, the process of determining the focusing depth, is as follows: for the main detection area, the geometric center point of the thickness layer is taken as the focusing point. This allows the main ultrasonic beam energy to uniformly cover the thickness of the layer, improving the detection sensitivity of defects within the layer.

[0052] For a type of secondary inspection area, the focal point is a point on the inner surface of the bolt hole that points radially toward the center of the hole. This concentrates the energy of the main ultrasonic beam on the inner surface of the bolt hole where defects are most likely to occur, which is beneficial for point defects or short cracks on the surface.

[0053] For the second-class secondary inspection area, the midpoint of the line connecting the centers of adjacent bolt holes is used as the center point in the thickness direction of the flange to be inspected. This concentrates the energy of the main ultrasonic beam in the middle of the second-class secondary inspection area, thereby effectively covering any possible extended cracks or band-shaped defects that penetrate the thickness.

[0054] A scanning coordinate system is established with the geometric center of the flange end face to be inspected as the origin, the two mutually perpendicular directions in the plane as the X-axis and Y-axis, and the direction perpendicular to the end face as the Z-axis.

[0055] Then, the path distance from the probe detection surface to each focal point along the direction of the corresponding main ultrasonic beam is calculated as the focusing depth.

[0056] The specific formula for calculating the sound path distance is as follows: .

[0057] in, The coordinates of the center of the probe array element; These are the coordinates of the focal point.

[0058] The center coordinates of the probe array elements are obtained by aligning the probe with the scanning coordinate system during probe calibration; the focal point coordinates are extracted based on the CAD model of the flange under inspection or the 3D scanning point cloud data.

[0059] In practical applications, if the flange to be inspected has abrupt changes in thickness, steps, bevels, or other structural features, an auxiliary focusing depth is added to the corresponding areas, such as the step height or the bevel root depth. This auxiliary focusing depth can be quantitatively determined through extraction from a CAD model or 3D scanning.

[0060] Finally, all focusing depths and auxiliary focusing depths are integrated to form a focusing depth set, which is then associated and paired with the incident angle set.

[0061] Please see Figure 4 Step S12, the process of determining the array element delay sequence is as follows: First, based on each incident angle and its associated pairing focusing depth, combined with the transverse wave sound velocity and the spatial position coordinates of each array element, the propagation time of the ultrasonic waves emitted by each array element to reach the corresponding focusing point synchronously is calculated.

[0062] The specific formula for calculating propagation time is as follows: .

[0063] in, For the first The ultrasonic waves emitted by each element reach the focal point. The spread time. For the first The spatial coordinates of each array element. The coordinates of the focal point; is the array element number, with a value range of 0, 1, 2, ..., k-1; k is the total number of array elements of the phased array probe.

[0064] Then, using the element with the longest propagation time as a reference, the time difference of the remaining elements relative to this reference is calculated as the transmission delay value. All transmission delay values ​​are arranged according to the element number to form an element delay sequence that uniquely corresponds to the current incident angle and focusing depth.

[0065] The probe emits according to the corresponding array element delay sequence in the main detection area, the first-class sub-detection area, and the second-class sub-detection area to achieve dynamic focusing in multiple areas.

[0066] S2. Perform multi-channel scanning according to the determined incident angle, focusing depth and array element delay sequence to obtain the ultrasound signal dataset.

[0067] The specific process is as follows: First, considering that the inspection surface of a wind turbine flange is usually the upper or lower end face, i.e., the flange face connected to the tower, the phased array probe is arranged circumferentially along the end face of the flange to be inspected. This circumferential arrangement facilitates a comprehensive scan of the entire annular flange surface.

[0068] Of course, depending on the actual structure of the flange to be inspected, a side-mounted arrangement can also be used. For example, at the flange neck or transition arc, the probe can be arranged circumferentially or axially along the side.

[0069] Then, based on the center coordinates and diameter of the bolt holes, a sequence of regular scanning points covering the entire inspection surface of the flange to be inspected is generated in the scanning coordinate system. The entire inspection surface refers to the entire annular end face excluding the area inside the bolt holes.

[0070] The specific process for generating a regular scan point sequence can be as follows: Using the outer diameter of the flange to be inspected as the outer boundary and the inner diameter as the inner boundary, a ring-shaped region is constructed within its end face plane as the scan boundary. Then, candidate scan points are generated within the boundary using a polar coordinate grid with equal spacing. Next, it is calculated whether the distance from each scan point to the center of each bolt hole is less than the hole radius; if so, the scan point is discarded. Finally, the remaining scan points constitute a regular scan point sequence covering the entire inspection surface, with the scan points evenly distributed according to a preset step size.

[0071] The step size should not exceed half the width of the effective ultrasonic main beam of the probe, typically 0.5 to 0.7 times the wafer size, to ensure sufficient overlap of the ultrasonic main beam between adjacent scanning points and avoid missed detections. In this invention, 0.6 times can be used as an example.

[0072] Furthermore, the incident angle set, focusing depth set, and array element delay sequence are loaded into the phased array controller. At each scanning point, each incident angle in the incident angle set and its associated paired focusing depth and transmission delay value are sequentially called to drive the phased array probe to perform multi-angle transmission and reception in a time-division manner.

[0073] Time-division refers to the process where, at the same scanning point, the probe emits ultrasonic waves sequentially with different incident angles, corresponding focusing depths, and emission delay values, and then receives the echoes.

[0074] During each ultrasonic wave transmission and echo reception process, the probe excites the corresponding channels according to the array element delay sequence to form an ultrasonic beam with a specific incident angle and focusing depth, and simultaneously acquires the ultrasonic echo signals of each channel.

[0075] After scanning all incident angles in the incident angle set at the current scanning point, the probe can be automatically controlled by the mechanical scanner to move to the next scanning point, repeating the multi-angle transmission and reception process until the scanning of the entire detection surface is completed.

[0076] The ultrasound echo signals of all channels and incident angles corresponding to each scanning point are integrated to form an ultrasound signal dataset containing scanning point coordinates, incident angles, channel numbers, and signal amplitudes.

[0077] Step S3: Extract the echo amplitude of each scanning point in the ultrasound signal dataset and compare it with the reference amplitude to obtain the amplitude deviation.

[0078] The specific process is as follows: Ultrasonic echo signals corresponding to each scanning point, incident angle, and channel are read from the ultrasonic signal dataset. First, the ultrasonic echo signals are rectified using a full-wave rectifier, and then the cutoff frequency is set to the probe's center frequency. to A low-pass filter is used for smoothing to obtain the signal envelope.

[0079] Then, within the preset depth analysis window, the peak value of the envelope is extracted as the echo amplitude of the scan point.

[0080] The center of the depth analysis window is set to the current focusing depth, and the window width is 1.5 to 2 times the width of the main ultrasonic beam at that focusing depth, so as to ensure coverage of the focusing area and suppress sidelobe interference.

[0081] Then, the echo amplitude is algebraically subtracted from the reference amplitude at the corresponding incident angle to obtain the amplitude deviation.

[0082] The reference amplitude can be determined as follows: a standard wind turbine flange test block made of the same batch of materials and heat treatment process as the flange under test, and confirmed to be free of internal and surface defects by multiple methods such as X-ray inspection and penetrant testing. Multi-channel ultrasonic testing is performed using the same incident angle set, focusing depth set, array element delay sequence, and scanning path as the flange under test.

[0083] Then, the ultrasonic echo signals corresponding to each scanning point, incident angle, and channel of the standard test block are extracted, and the statistical average value of the echo amplitude is calculated as the reference amplitude under the testing conditions. The reference amplitude is used to establish a three-dimensional mapping relationship based on the scanning coordinates, incident angle, and channel number, and is stored.

[0084] Please see Figure 5 Step S4: Dynamically identify and output a set of abnormal regions based on amplitude deviation and geometric dimension data.

[0085] The specific process is as follows: For each scanning point, select the one with the largest absolute value of the amplitude deviation between all incident angles and the channel as the comprehensive amplitude deviation.

[0086] Scan points whose absolute value of the overall amplitude deviation exceeds a preset threshold are then selected as outliers.

[0087] Since the amplitude of defect signals is usually significantly higher than that of background noise, setting the threshold too high may lead to missed detections, while setting it too low may easily introduce false alarms. Therefore, it is necessary to balance the risks of missed detections and false alarms. In this invention, three times the standard deviation of the reference amplitude is exemplarily taken as the preset threshold.

[0088] For scenarios requiring high detection rates, the standard deviation can be appropriately reduced to 2 times the baseline amplitude standard deviation; for scenarios requiring low false alarm rates, the standard deviation can be increased to 4 times the baseline amplitude standard deviation to adapt to different detection needs.

[0089] Considering that defects typically have spatial continuity, the anomalous signals they generate will appear at adjacent scan points. Conversely, isolated, discrete anomalous points are more likely to be caused by noise or coupling fluctuations.

[0090] Therefore, spatial clustering of outliers is performed to merge multiple outlier signal points representing the same defect, thereby more accurately describing the geometric extent of the defect.

[0091] Spatial clustering algorithms are not limited to DBSCAN algorithm, K-means algorithm, etc., and this invention adopts DBSCAN algorithm.

[0092] The clustering process is as follows: Starting from any outlier, find all outliers whose spatial distance is less than a set connection distance, and group them into the same set. Then, recursively expand this process until no new outliers can be added, ultimately forming an outlier region. Repeat this process until all outliers have been visited.

[0093] The connection distance can be determined based on the probe's scanning step size. In this invention, 1.5 times the scanning step size is used as the set connection distance to ensure that adjacent anomalies belonging to the same defect are correctly clustered. For cases where the expected defect size is small or fine segmentation is required, the set connection distance can be reduced to 1 times the scanning step size.

[0094] Subsequently, the average coordinates of all points within the abnormal region are calculated as the center position. Simultaneously, the maximum coordinate difference between the abnormal region and the X and Y directions in the scanning coordinate system is calculated, and the maximum value is taken as the spatial extension length, thus reflecting the maximum span of the defect projected onto the detection surface.

[0095] Furthermore, the maximum absolute value among all amplitude deviations within the abnormal region is extracted as the maximum amplitude deviation intensity. The maximum amplitude deviation intensity represents the maximum signal response intensity that the defect may produce, reflecting the severity of the defect.

[0096] Finally, the output includes a set of anomalous regions, including the center location, spatial extension length, and maximum magnitude deviation intensity.

[0097] Step S5: Determine the defect type based on the center location, spatial extension length, maximum deviation intensity, and geometric dimensions of each abnormal area.

[0098] First, determine the structural region type to which the abnormal region belongs based on the coordinates of the center position of each abnormal region and the center position of the bolt hole.

[0099] The specific method is to calculate the first distance from the center of the abnormal area to the center of the nearest bolt hole, and the second distance to the line connecting the centers of the nearest adjacent bolt holes.

[0100] If the first distance is less than or equal to the set radial distance corresponding to a sub-detection area, then the abnormal area is determined to be the area adjacent to the bolt hole. Otherwise, proceed to the next judgment.

[0101] The area adjacent to a bolt hole refers to a circular region centered on the bolt hole and with a radius equal to the sum of the bolt hole radius and the extended distance. This area experiences localized stress concentration and is prone to radial or circumferential point defects, short cracks, or clusters of micropores.

[0102] Considering that defects are usually concentrated at the edge of the bolt hole, the extension distance can be exemplarily taken as 0.5 times the bolt hole diameter.

[0103] If the second distance is less than or equal to the set width corresponding to the secondary detection area, the abnormal area is determined to be the area between adjacent bolt holes. Otherwise, it is determined to be the flange body area.

[0104] The area between adjacent bolt holes refers to a rectangular or strip-shaped region extending to both sides of a predetermined width, based on the line connecting the centers of each pair of adjacent bolt holes. This area is subjected to combined shear and tensile stresses, making it prone to developing extended cracks, band-shaped defects, or incomplete fusion that propagate along the hole-to-hole direction.

[0105] In the scanning coordinate system, the remaining area after removing all areas adjacent to bolt holes and the areas between adjacent bolt holes on the inspection surface of the flange is the flange body area. The flange body area is prone to planar defects due to overall bending and axial load.

[0106] After determining the structural region type, the defect type is then determined by combining the spatial extension length of the abnormal region.

[0107] The specific determination process is as follows: If the abnormal area is located in the vicinity of the bolt hole, and its spatial extension length is less than the bolt hole diameter, it is determined to be a point defect at the hole edge, such as casting porosity, micro-inclusions, machining burrs, or other small isolated defects. Otherwise, it is determined to be an extended defect at the hole edge.

[0108] If the abnormal area belongs to the region between adjacent bolt holes, and its spatial extension length is greater than the center-to-center distance between adjacent bolt holes, it suggests the existence of an extended crack or band-like defect connecting the two bolt holes, and is therefore determined to be an inter-hole extended defect. Otherwise, it is determined to be an inter-hole local defect, such as point inclusions or short cracks.

[0109] If the abnormal area belongs to the flange body area, and its spatial extension length is greater than the bolt hole diameter but less than the center distance between adjacent bolt holes, it indicates that there may be large delamination, flaky inclusions, or forging folds and other planar defects. It is then determined to be a planar defect of the body. Planar defects of the body will affect the overall load-bearing capacity and fatigue life of the flange under inspection.

[0110] If the spatial extension length of the flange body area to which the abnormal area belongs is less than or equal to the bolt hole diameter, it is determined to be a point defect in the body, such as porosity or inclusions; if the spatial extension length is greater than or equal to the center distance between adjacent bolt holes, it is determined to be a large cross-regional defect, such as large-scale delamination or cross-hole cracks.

[0111] Finally, the determination result is output as the defect type and associated with the abnormal area.

[0112] Step S6: Associate and store the geometric dimensions, incident angle, focusing depth, defect type and abnormal area set of this inspection, and periodically optimize and correct the incident angle and focusing depth based on the statistical analysis of historical associated and stored data.

[0113] The specific process is as follows: after every 100 tests or once a month, records with the largest deviation intensity greater than the preset intensity threshold in the historical associated storage data can be periodically extracted.

[0114] The preset intensity threshold can be taken as the 80th percentile of the maximum magnitude deviation intensity in the historical associated stored data. This eliminates most noise or minor anomalies, retains obvious defect signals, and ensures sufficient positive samples for statistical analysis. For the first inspection, the preset intensity threshold can be the statistical value of similar flange inspection data.

[0115] Each record includes geometric dimensions, incident angle, focusing depth, defect type, center location of the abnormal area, and maximum deviation intensity.

[0116] Then, because the optimal incident angle and focusing depth differ for different types of defects in different regions, it is necessary to first group the records according to defect type and structural region type. Then, the average value of the maximum deviation intensity corresponding to each combination of incident angle and focusing depth in each group, as well as the detection frequency of that combination, are calculated.

[0117] Among them, the detection frequency refers to the number of times that a certain combination of incident angle and focusing depth successfully detects the same type of defect.

[0118] Subsequently, the combination of incident angle and focusing depth with the largest average value and a detection frequency exceeding the set number is selected as the optimization parameter for this defect type under the corresponding structural region type.

[0119] The maximum average value indicates that the combination of incident angle and depth of focus has historically produced the strongest signal response to this type of defect.

[0120] The number of optimization attempts can be set based on the amount of historical data. For example, if there is sufficient data after grouping, at least 3 or 5 attempts can be required. If the data volume is insufficient, optimization can be temporarily suspended, and the default parameters can be used.

[0121] If there are no defect samples in the historical associated storage data that meet the criteria of having the largest average value and a detection frequency exceeding the set number, the current default parameters will continue to be used, and optimization correction will not be performed temporarily to avoid parameter overfitting or misadjustment due to insufficient samples.

[0122] When inspecting flanges of the same geometric dimensions in subsequent tests, the structural region type is identified based on the geometric dimension data of the flange to be inspected, and the corresponding optimization parameters are loaded into the phased array controller for initial scanning.

[0123] After testing is completed, the results are updated to the historical data for optimization and correction in the next cycle. This reduces reliance on general parameters, improves the detection rate and efficiency of initial testing, and adapts to rapid batch testing. Simultaneously, continuously feeding new test results back to the database allows for adaptation to fluctuations in on-site processes and differences in material batches, achieving dynamic updates and continuous improvement to meet the rapid and adaptive testing needs of industrial production lines.

[0124] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0125] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0126] In addition, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0127] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0128] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges, characterized in that, include: S1. Based on the geometric dimensions and material grade of the flange to be inspected, determine the incident angle, focusing depth, and array element delay sequence for ultrasonic testing; S2. Perform multi-channel scanning according to the determined incident angle, focusing depth and array element delay sequence to acquire ultrasound signal dataset; S3. Extract the echo amplitude of each scanning point in the ultrasound signal dataset and compare it with the reference amplitude to obtain the amplitude deviation. S4. Dynamically identify and output a set of abnormal regions based on amplitude deviation and geometric dimension data; the set of abnormal regions includes at least the center position, spatial extension length and maximum amplitude deviation intensity of each abnormal region. S5. Determine the defect type based on the center location, spatial extension length, maximum deviation intensity, and geometric dimensions of each abnormal area; S6. Associate and store the geometric dimensions, incident angle, focusing depth, defect type and abnormal area set of this inspection, and periodically optimize and correct the incident angle and focusing depth based on the statistical analysis of historical associated and stored data; The process of determining the incident angle is as follows: the flange to be inspected is divided into several continuous and adjacent main detection areas along the thickness direction, and the direction of the main ultrasonic beam extends along the thickness direction; an annular area formed by extending a set radial distance outward with the center of the bolt hole as the center is used as a first-class secondary detection area, and the direction of the main ultrasonic beam extends radially along the flange to be inspected and points to the central axis of the bolt hole at the corresponding scanning point; a strip-shaped area formed by extending a set width to both sides with the line connecting the centers of adjacent bolt holes as the reference is used as a second-class secondary detection area, and the direction of the main ultrasonic beam is perpendicular to the line connecting the centers of adjacent bolt holes. Based on the transverse wave velocity and according to the acoustic refraction law, the incident angles of the main detection area, the first-class secondary detection area, and the second-class secondary detection area are calculated respectively to form a set of incident angles.

2. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 1, characterized in that, The geometric dimensions include the thickness of the flange to be inspected, the center coordinates and diameter of the bolt holes; and the corresponding transverse wave velocity is determined according to the material grade.

3. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 1, characterized in that, The process for determining the focusing depth is as follows: The geometric center point of the main inspection area in the thickness direction of the flange under inspection is taken as the focal point; For a type of secondary detection area, the point on the inner surface of the bolt hole pointing radially toward the center of the hole is taken as the focal point; For the second-class secondary inspection area, the focal point is the center of the line connecting the centers of adjacent bolt holes in the thickness direction of the flange to be inspected; The path distance from the probe detection surface to each focal point along the direction of the corresponding main ultrasonic beam is calculated as the focal depth, forming a focal depth set, which is then associated and paired with the incident angle set.

4. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 2, characterized in that, The process for determining the array element delay sequence is as follows: Based on each incident angle and its associated pairing focusing depth, combined with the transverse wave velocity and the spatial position coordinates of each array element, the propagation time of the ultrasonic waves emitted by each array element to reach the corresponding focusing point synchronously is calculated. Using the array element with the longest propagation time as the benchmark, the time difference of the remaining array elements relative to the benchmark is calculated as the transmission delay value. All emission delay values ​​are arranged according to the array element number to form an array element delay sequence that uniquely corresponds to the current incident angle and focusing depth.

5. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 1, characterized in that, Step S2 specifically involves: Arrange the probe around the end face of the flange to be inspected, and establish a scanning coordinate system with the geometric center of the flange to be inspected as the origin; Load the incident angle set, focusing depth set, and array element delay sequence into the phased array controller; Based on the center coordinates and diameter of the bolt holes, a regular sequence of scanning points covering the entire inspection surface of the flange to be inspected is generated in the scanning coordinate system. At each scanning point, the incident angles in the incident angle set and their associated paired focusing depths and transmission delay values ​​are called sequentially to drive the phased array probe to perform multi-angle transmission and reception in a time-division manner. During each transmission, the probe excites the corresponding channel according to the array element delay sequence to form an ultrasonic beam with a specific incident angle and focusing depth, and simultaneously acquires the ultrasonic echo signal of each channel; After the current scanning point has completed scanning of all incident angles, the probe is moved to the next scanning point, and the multi-angle transmission and reception process is repeated until the scanning of the entire detection surface is completed; The ultrasound echo signals of all channels and incident angles corresponding to each scanning point are integrated to form an ultrasound signal dataset containing scanning point coordinates, incident angles, channel numbers, and signal amplitudes.

6. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 1, characterized in that, Step S3 specifically involves: Read the ultrasound echo signals corresponding to each scanning point, each incident angle, and each channel from the ultrasound signal dataset; Envelope detection processing is performed on the ultrasonic echo signal, and the envelope peak value is extracted as the echo amplitude. The amplitude deviation is obtained by algebraically subtracting the echo amplitude from the reference amplitude at the corresponding incident angle.

7. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 2, characterized in that, The process of dynamically identifying and outputting the set of abnormal regions is as follows: The maximum absolute value of the amplitude deviation between all incident angles corresponding to each scanning point and the channel is selected as the comprehensive amplitude deviation. Scan points whose absolute value of the overall amplitude deviation exceeds a preset threshold are selected as outliers; Spatial clustering of outliers is performed, and outliers whose spatial distance is less than a set connection distance are aggregated into outlier regions. Calculate the average coordinates of all points within the abnormal region as the center location; Calculate the maximum coordinate difference of the abnormal region in each dimension of the scanning coordinate system, and take the maximum value as the spatial extension length; Extract the maximum absolute value among all amplitude deviations within the abnormal region as the intensity of the maximum amplitude deviation; The output includes a set of anomalous regions with center location, spatial extension length, and maximum magnitude deviation intensity.

8. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 7, characterized in that, Step S5 specifically involves: Based on the coordinates of the center position of each abnormal area and the center position of the bolt hole, determine the structural area type to which the abnormal area belongs; The structural region types include bolt hole adjacent region, region between adjacent bolt holes, or flange body region. If the defect is located in the vicinity of the bolt hole and its spatial extension length is less than the bolt hole diameter, it is determined to be a point defect at the hole edge; otherwise, it is determined to be an extended defect at the hole edge. If the area between adjacent bolt holes has a spatial extension length greater than the center-to-center distance between adjacent bolt holes, it is determined to be an inter-hole extension type defect; otherwise, it is determined to be an inter-hole local defect. If the defect is located within the flange body area and its spatial extension length is greater than the bolt hole diameter but less than the center-to-center distance between adjacent bolt holes, it is determined to be a surface defect of the body; otherwise, it is determined to be a point defect of the body or a large defect spanning multiple areas. Output the defect type and associate it with the abnormal region.

9. The multi-channel ultrasonic non-destructive intelligent testing method for wind turbine flanges according to claim 1, characterized in that, Step S6 specifically involves: Periodically extract records from historical associated storage data where the largest deviation intensity exceeds a preset intensity threshold; Each record includes geometric dimensions, incident angle, depth of focus, defect type, center location of the abnormal area, and maximum deviation intensity. The records were grouped according to defect type and structural region type; the average value of the maximum deviation intensity corresponding to each combination of incident angle and focusing depth in each group and the detection frequency of that combination were calculated. The combination of incident angle and focusing depth with the largest average value and a detection frequency exceeding a set number is selected as the optimization parameter; When inspecting flanges of the same geometric dimensions in subsequent tests, the structural region type is identified based on the geometric dimension data of the flange to be inspected, and the corresponding optimization parameters are loaded into the phased array controller for initial scanning. After the test is completed, the test results will be updated to the historical associated storage data for optimization and correction in the next cycle.

Images