Weld defect detection method, device and equipment based on three-dimensional holographic scanning

Abstract

The invention relates to the technical field of welding seam defect detection, and discloses a welding seam defect detection method, device and equipment based on three-dimensional holographic scanning. Coherent light is used for irradiating a welding seam to be detected, a reference state hologram sequence and a loading state hologram sequence are generated, a phase difference field is calculated, and the deformation gradient is analyzed; marking a first abnormal region and extracting a deformation concentration ratio and a spatial distribution mode, calculating a dominant spatial frequency through space-frequency domain joint analysis, fusing abnormal features, marking a second abnormal region and extracting amplitude distribution and spatial continuity parameters, and performing defect characterization analysis by combining information of the first and second abnormal regions to obtain weld defect detection information. The method has the advantages of high precision, non-contact and real-time monitoring, the accuracy and efficiency of weld defect detection can be effectively improved, and the problem that weld detection is not accurate enough in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of weld defect detection, and in particular to a weld defect detection method, device and equipment based on three-dimensional holographic scanning. Background Art

[0002] Welds are important structural parts that connect two metal components. Their quality is directly related to the strength, durability and safety of the product. In industrial manufacturing, the presence of welding defects may lead to structural fragility, especially under environmental conditions such as dynamic loads and thermal stress. Weld defects will be magnified and may even cause equipment failure. Therefore, accurate and rapid weld defect detection has become an important link to ensure welding quality and improve project safety. Traditional weld defect detection methods (such as visual inspection, X-rays, ultrasound, etc.) have achieved success to a certain extent, but these methods have some limitations. The detection of complex weld shapes is not accurate, the detection process is cumbersome, and the experience of the inspectors is required to be high. In addition, traditional detection methods mostly rely on manual operation, making it difficult to achieve fully automatic and efficient weld defect monitoring. Summary of the Invention

[0003] The purpose of the present invention is to provide a weld defect detection method, device and equipment based on three-dimensional holographic scanning, aiming to solve the problem of insufficient accuracy of weld detection in the prior art.

[0004] The present invention is implemented as follows: In a first aspect, the present invention provides a weld defect detection method based on three-dimensional holographic scanning, comprising: The weld to be inspected is irradiated with coherent light in a sequence of no load and dynamic environmental load to generate a reference state hologram and a loaded state hologram sequence; Calculating each phase difference field between the loaded state hologram sequence and the reference state hologram, and calculating the local deformation gradient of each phase difference field to mark the first abnormal area of ​​the weld to be inspected and obtain the corresponding deformation concentration and spatial distribution pattern; The dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram are calculated by joint analysis in the space-frequency domain to fuse the abnormal features of each dominant spatial frequency, so as to mark the second abnormal area of ​​the weld to be inspected and obtain the corresponding amplitude distribution and spatial continuity parameters; Based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, as well as the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameters, a defect characterization analysis is performed on the weld to be inspected to obtain weld defect detection information.

[0005] In a second aspect, the present invention provides a weld defect detection device based on three-dimensional holographic scanning, which is used to implement the weld defect detection method based on three-dimensional holographic scanning described in any one of the first aspects, comprising: A dynamic testing module is used to irradiate the weld to be inspected with coherent light in a sequence of unloaded and dynamic environmental loads to generate a reference state hologram and a loaded state hologram sequence; a phase analysis module, configured to calculate each phase difference field between the loaded state hologram sequence and the reference state hologram, and to calculate a local deformation gradient for each phase difference field, so as to mark a first abnormal area of ​​the weld to be inspected and obtain a corresponding deformation concentration and spatial distribution pattern; a frequency analysis module for calculating the dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram through joint analysis in the space-frequency domain, so as to fuse the abnormal features of each dominant spatial frequency, mark the second abnormal area of ​​the weld to be inspected, and obtain the corresponding amplitude distribution and spatial continuity parameters; The defect analysis module is used to perform defect characterization analysis on the weld to be inspected based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, as well as the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameters, so as to obtain weld defect detection information.

[0006] In a third aspect, the present invention provides a weld defect detection device based on three-dimensional holographic scanning, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements a weld defect detection method based on three-dimensional holographic scanning as described in any one of the first aspects.

[0007] The present invention provides a weld defect detection method based on three-dimensional holographic scanning, which has the following beneficial effects: The present invention uses coherent light to illuminate the weld to be inspected, generates a reference state and a loaded state hologram sequence, calculates the phase difference field and analyzes the deformation gradient, marks the first abnormal area and extracts the deformation concentration and spatial distribution pattern, calculates the dominant spatial frequency through joint analysis of the space-frequency domain, fuses the abnormal features, marks the second abnormal area and extracts the amplitude distribution and spatial continuity parameters, combines the information of the first and second abnormal areas to perform defect characterization analysis, and obtains weld defect detection information. This method has the advantages of high precision, non-contact and real-time monitoring, can effectively improve the accuracy and efficiency of weld defect detection, and solves the problem of insufficient accuracy of weld detection in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 1 is a schematic diagram of the steps of a weld defect detection method based on three-dimensional holographic scanning provided by an embodiment of the present invention; Figure 2It is a structural schematic diagram of a weld defect detection device based on three-dimensional holographic scanning provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0009] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0010] The implementation of the present invention is described in detail below with reference to specific embodiments.

[0011] Reference Figure 1 、 Figure 2 As shown, a preferred embodiment of the present invention is provided.

[0012] In a first aspect, the present invention provides a weld defect detection method based on three-dimensional holographic scanning, comprising: S1: The weld to be inspected is illuminated with coherent light under unloaded and dynamic environmental load sequences to generate a reference state hologram and a loaded state hologram sequence; S2: Calculating each phase difference field between the loaded state hologram sequence and the reference state hologram, and calculating the local deformation gradient of each phase difference field to mark the first abnormal area of ​​the weld to be inspected and obtain the corresponding deformation concentration and spatial distribution pattern; S3: Calculating the dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram through joint analysis in the space-frequency domain to fuse abnormal features of each dominant spatial frequency, thereby marking the second abnormal area of ​​the weld to be inspected and obtaining the corresponding amplitude distribution and spatial continuity parameters; S4: Based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, as well as the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameters, a defect characterization analysis is performed on the weld to be inspected to obtain weld defect detection information.

[0013] Specifically, in step S1 of the embodiment provided by the present invention, the weld to be inspected is maintained in a stable state without applying any external load. This stable state refers to the original state of the weld when it is not subject to external forces such as heat, pressure or vibration. The weld in the stable state can be used as a reference state for comparison with the changes in the loaded state. Only under this condition can the holographic information of the weld itself be accurately captured and errors caused by external influences be avoided. By ensuring the stable state of the weld, external interference can be effectively reduced, ensuring that the generated reference state hologram truly reflects the morphological characteristics of the weld under no load.

[0014] More specifically, a pulsed laser emits coherent light at the weld to be inspected in a stable state. A beam splitter separates the coherent light into object light and reference light. The object light impinges on the weld and is reflected before reaching the camera, while the reference light directly reaches the camera's target surface. The object and reference lights interfere on the camera's target surface, forming a reference-state hologram. Coherent light is used to ensure the interference effect of light waves, thereby obtaining a high-resolution hologram. This interference-formed hologram provides detailed information on the weld's surface and internal structure. The resulting reference-state hologram can very accurately record the weld's three-dimensional morphology, surface microscopic features, and any potential defects in its stable state.

[0015] More specifically, microsecond thermal pulses and low-frequency mechanical vibrations that meet the environmental load specifications are applied to the weld to be inspected, causing the weld to enter a deformation excitation state. These loads simulate the temperature changes and vibration effects in the actual use environment. By applying dynamic loads, the deformation and stress that may occur in the weld in actual applications can be stimulated, simulating the actual load conditions of the weld during use, which helps to reveal defects or stress concentration areas that are not detectable in a static state. When the weld is in a deformation excitation state, its surface and interior will deform, and these deformations will affect the interference pattern of the hologram. Through this process, subtle changes in the weld under loading can be revealed, providing key information for subsequent defect detection.

[0016] More specifically, after applying the dynamic environmental load, the pulsed laser is continued to emit coherent light to the weld, irradiating the weld and recording the loaded state hologram. Under the loaded state, the morphology and structure of the weld may change, and tiny cracks, expansion, deformation and other phenomena may occur. By obtaining the loaded state hologram, these changes can be captured and compared with the reference state hologram for analysis. By recording the loaded state hologram, the three-dimensional morphological characteristics of the weld under stress can be obtained. Compared with the reference state hologram, the difference in the loaded state hologram reveals the deformation of the weld and possible defect areas.

[0017] More specifically, based on the dynamic environmental load sequence, the load applied to the weld (such as the intensity and frequency of thermal pulses and mechanical vibrations) is continuously adjusted so that the weld is in different load states in turn, thereby generating a series of loaded holograms. Under different load conditions, the weld will exhibit different deformation modes. By recording and analyzing the holograms under multiple load states, the deformation and defects of the weld can be more comprehensively reflected. By collecting the hologram sequence, the dynamic response of the weld under multiple loads can be obtained, which provides a rich data source for subsequent phase difference calculation, deformation gradient analysis and defect detection. This process uses high-precision holographic scanning technology to generate fine reference state and loaded state holograms, capture the tiny deformation of the weld, and provide accurate defect detection data. Through the hologram sequence under multiple dynamic load conditions, the behavior of the weld in different working states can be fully reflected, helping to reveal potential defects. Through coherent light irradiation and phase difference analysis, high-quality weld defect detection results can be obtained, reducing errors and false positives and improving detection reliability.

[0018] Specifically, in step S2 of the embodiment provided by the present invention, a reference state hologram and a loaded state hologram are obtained, and the two are reconstructed using the angular spectrum method to obtain a complex amplitude. Then, the phase difference field of the two holograms is calculated. The phase difference is the core indicator of the change in the hologram under the loaded state. It reflects the slight deformation of the weld morphology caused by the external load. The calculation of the phase difference can directly reveal the changes in the weld surface morphology and deformation. By calculating the phase difference field, the change of the weld between the loaded state and the reference state can be quantified. This change can reflect the deformation of the weld under dynamic load, such as deformation, cracks, or other defects, with high resolution and meticulous measurement accuracy.

[0019] More specifically, based on the phase difference field, a gradient operator (such as the Sobel operator or the central difference method) is used to calculate the spatial gradient of the displacement field, and then the shear strain and normal strain are calculated based on the displacement field. The deformation gradient can reflect the local deformation characteristics of the weld during loading, especially in areas where stress is concentrated or cracks are formed. Through gradient calculation, the strain distribution pattern of the local area can be accurately identified. Through deformation gradient calculation, the stress distribution in the weld can be revealed, and the area with the most concentrated strain can be marked. This is key data for subsequent defect analysis and evaluation of the weld structural strength.

[0020] More specifically, based on the deformation gradient results, the areas of greatest deformation concentration are identified. These areas typically exhibit high strain values ​​(e.g., deformation gradients greater than a preset threshold). Using image segmentation algorithms or thresholding (e.g., Canny edge detection), high-deformation regions are extracted. These regions are often associated with defects in the weld, such as cracks, holes, or stress concentrations. Marking these areas helps quickly locate potential weld problems and identifies the first abnormal areas, providing key areas of focus for defect analysis. These areas may represent weak links in the weld, or areas at high risk of fracture, cracks, or other problems.

[0021] More specifically, the deformation concentration of the marked first abnormal area is calculated. Deformation concentration is an indicator of the deformation gradient and is usually calculated as the ratio of the maximum deformation gradient to the average deformation gradient in the area. At the same time, through spatial distribution analysis, the spatial continuity of these high-deformation areas can be analyzed. Cluster analysis or geometric analysis methods can be used to identify the morphological characteristics of the abnormal area, such as crack propagation paths or stress concentration points. The deformation concentration and spatial distribution pattern can reveal the severity and spatial characteristics of the weld defects and help determine the location, size and possible impact range of the defects. The calculation of deformation concentration and spatial distribution pattern can help locate and evaluate defects in the weld and further analyze their impact on the overall structure of the weld. This provides specific guidance for subsequent repair work.

[0022] It can be understood that through the calculation of phase difference and deformation gradient, tiny deformations and defects in the weld can be revealed with high precision, and the problem area can be located. Through the analysis of deformation concentration and spatial distribution pattern, a comprehensive description of weld defects can be obtained, including the severity and distribution of the defects. The phase difference field, deformation gradient and abnormal area marking images generated by the whole process can intuitively display weld defects and provide a clear defect detection report.

[0023] Specifically, in step S3 of the embodiment provided by the present invention, the reference state hologram and the loaded state hologram are converted from the spatial domain to the frequency domain, which is usually achieved using Fourier transform. This process can convert the spatial information in the hologram into frequency information, thereby revealing the contribution of different spatial frequency components in the image. The detailed information in the hologram (such as tiny cracks or deformations) may be difficult to observe directly in the spatial domain. By converting to the frequency domain, especially focusing on the dominant spatial frequency, the frequency characteristics related to the defect can be better identified. The frequency domain transformation can separate the high-frequency components (representing subtle changes) and low-frequency components (representing large-scale features) in the image, making the subtle changes in the defect area more prominent and easy to analyze.

[0024] More specifically, the frequency domain information obtained through Fourier transform is used to extract the dominant spatial frequencies in the hologram. Dominant spatial frequencies refer to frequency components with large amplitudes in the frequency domain. These components are often closely related to image details (such as the microscopic morphology of defects). The amplitude spectrum of the dominant spatial frequencies is calculated, and the dominant frequencies are determined based on the distribution of the amplitude spectrum. The amplitude of the frequency component can be calculated using the following formula. The dominant spatial frequency represents the most prominent detail features in the image and can effectively distinguish normal from abnormal areas of the weld. By extracting these dominant frequency components, defects and stress concentration areas in the weld can be accurately identified. After calculating the dominant spatial frequency, the characteristics of the weld defects in the frequency domain can be determined. The high-frequency portion is often associated with subtle deformations such as cracks and stress concentrations in the weld, effectively revealing subtle weld defects.

[0025] More specifically, the amplitudes of the dominant spatial frequencies in the reference state and loaded state holograms are fused and analyzed, and the difference between the two is detected in combination with the spatial frequency distribution. The frequency difference diagram (obtained from the dominant frequency difference between the reference state and the loaded state) is used to mark abnormal areas in the weld. By setting a threshold, areas with large amplitude differences can be marked. These areas are usually related to weld defects (such as cracks, voids, etc.). By fusing the frequency characteristics of the loaded state and the reference state, the deformation and defects generated during the loading process can be further highlighted. The difference in the frequency domain can effectively reveal the abnormal response of the weld under dynamic loads. Through frequency feature fusion, the abnormal areas in the weld can be accurately marked, providing a direct basis for subsequent defect analysis. These abnormal areas are usually cracks, stress concentration areas or other potential defects in the weld.

[0026] More specifically, after marking the abnormal area, the severity and spatial distribution characteristics of the defect are further analyzed by calculating the amplitude distribution and spatial continuity parameters of the area. The amplitude distribution can be evaluated by calculating the statistical characteristics of the amplitude in the abnormal area (such as maximum, minimum, average and standard deviation, etc.), and the spatial continuity parameters can be obtained by analyzing the spatial uniformity of the frequency distribution in the abnormal area. If the frequency distribution of the area changes dramatically, it means that there may be sudden deformation or cracks in the area. The amplitude distribution and spatial continuity parameters can reveal the specific morphology of the defect area and its spatial expansion, which is crucial for judging the impact range of the defect and possible structural damage. Through amplitude distribution and spatial continuity analysis, the spatial characteristics of the defect, such as the morphology, distribution range and potential expansion path of the defect, can be accurately described, which provides detailed defect data and helps subsequent repair and maintenance work.

[0027] More specifically, after the fusion analysis, the second abnormal area of ​​the weld is determined by calculating the amplitude distribution and spatial continuity parameters. These abnormal areas usually correspond to more serious defects or potential failure areas in the weld. The locality and continuity of the frequency difference can be further analyzed to determine the expansion path of the defect and the possible crack propagation trend. The second abnormal area usually refers to the more serious area caused by stress concentration or crack propagation during the loading process. These areas can be accurately identified and located through frequency difference and spatial characteristic analysis. The marked second abnormal area can provide a detailed defect analysis diagram to help locate key areas that may affect the safety of the weld and the stability of the structure.

[0028] It is understandable that frequency analysis can more precisely identify defective areas in welds, especially under loading. This method can reveal tiny deformations and anomalies. By fusing frequency characteristics and spatial continuity parameters, the nature, morphology and expansion trend of weld defects can be comprehensively evaluated, providing data support for subsequent processing. By marking abnormal areas and calculating amplitude distribution, a clear defect visualization map can be provided to effectively judge the quality and safety of welds. The joint analysis of space and frequency domains can not only accurately mark weld defects, but also reveal the detailed characteristics and spatial distribution of defects, providing an important basis for subsequent weld repair and quality control.

[0029] Specifically, in step S4 of the embodiment provided by the present invention, the deformation concentration and spatial distribution pattern of the first abnormal area in the weld are extracted by joint analysis in the space-frequency domain. The deformation concentration can be obtained by comparing the loaded state and reference state holograms in the frequency domain. The higher the concentration, the greater the deformation in the area, which may be a precursor to a defect. The spatial distribution pattern refers to the distribution characteristics of the deformation or abnormal signal in the area in space, which can be determined by analyzing the spatial locality of the frequency components in the area. The first abnormal area is usually the early deformation area of ​​the weld, which may be due to stress concentration caused by factors such as material inhomogeneity and temperature gradient. By analyzing the deformation concentration and spatial distribution of the area, potential defect areas can be identified and preliminary calibration can be performed. It can identify small deformation areas in the weld and provide early warnings for subsequent defect detection. The quantitative analysis of the deformation concentration helps to reveal the stress concentration areas of the weld and identify the occurrence of cracks or other defects in advance.

[0030] More specifically, the amplitude distribution and spatial continuity parameters of the second abnormal area are calculated. The amplitude distribution represents the vibration or deformation intensity in the area, and the spatial continuity parameters reveal the expansion trend of the defects in the area. When calculating the amplitude distribution, the amplitude spectrum in the frequency domain can be used to select the high-frequency part to observe the performance of small defects such as cracks. The spatial continuity parameters reveal the extensibility of the defects by analyzing the spatial uniformity or unevenness of the frequency components in the defect area. Areas with poor continuity may indicate the propagation of defects such as cracks in the area. The second abnormal area is usually a more serious or deep defect area. By analyzing the amplitude distribution and spatial continuity parameters, the morphology and expansion path of the defects can be more accurately depicted, especially large-scale cracks or material failure areas. Through amplitude distribution and spatial continuity analysis, the more serious defect areas in the weld can be effectively marked, providing an accurate basis for subsequent repair and quality assessment. This step provides more accurate information for further analysis of the defects.

[0031] More specifically, the characteristic data of the first abnormal area and the second abnormal area are integrated to analyze the type and severity of the weld defects, compare the deformation concentration, amplitude distribution, spatial distribution pattern and spatial continuity parameters of different abnormal areas, and comprehensively evaluate the structural characteristics of the weld defects. Based on the analysis results, combined with engineering experience or historical data, the possible types of weld defects (such as cracks, pores, welding defects, etc.) are judged, and finally the weld defect detection information is obtained to form comprehensive characterization data of the weld defects. The purpose of defect characterization analysis is to comprehensively evaluate the defect situation of the weld by integrating multi-dimensional characteristic data. By comparing the characteristics of different abnormal areas, various types of defects can be effectively distinguished and the severity of the defects can be quantified. This step can provide a comprehensive and accurate weld defect detection report describing the type, location and possible impact of the defect, which is of great significance for the subsequent detection, repair and quality control of the weld.

[0032] More specifically, based on the results of the defect characterization analysis, detection information of weld defects is generated, including the specific location, type, shape, size, expansion direction and possible affected area of ​​the defects. This detection information can generate a spatial distribution map of the defects through visualization tools to help intuitively understand the defect characteristics of the weld. If necessary, defect risk assessment information can also be provided, such as crack expansion trend, structural reliability assessment, etc. The purpose of generating defect detection information is to provide accurate decision-making basis for subsequent repairs, assessments and safety inspections. This information can help quickly locate problems and take measures. By generating detailed defect detection information, it can provide precise guidance for subsequent control and maintenance of weld quality, ensuring the safety and reliability of welds during use.

[0033] It can be understood that by analyzing the deformation concentration of the first abnormal zone, potential micro-deformations and stress concentration areas of the weld can be identified as early as possible. Through amplitude distribution and spatial continuity analysis, the morphology and expansion trend of serious defects such as cracks and pores in the weld can be more accurately characterized. Combining the analysis results of the first and second abnormal zones, comprehensive characterization information of weld defects can be generated to help decision makers make more reasonable judgments during the repair and quality control process, provide accurate data for subsequent weld inspection and repair, enhance the level of refinement of weld quality control, and effectively reduce the safety risks caused by defects in welds. Through the above steps, accurate analysis and characterization of weld defects can be achieved, which helps ensure weld quality, optimize production processes, and improve the safety and reliability of structures.

[0034] The present invention provides a weld defect detection method based on three-dimensional holographic scanning, which has the following beneficial effects: The present invention uses coherent light to illuminate the weld to be inspected, generates a reference state and a loaded state hologram sequence, calculates the phase difference field and analyzes the deformation gradient, marks the first abnormal area and extracts the deformation concentration and spatial distribution pattern, calculates the dominant spatial frequency through joint analysis of the space-frequency domain, fuses the abnormal features, marks the second abnormal area and extracts the amplitude distribution and spatial continuity parameters, combines the information of the first and second abnormal areas to perform defect characterization analysis, and obtains weld defect detection information. This method has the advantages of high precision, non-contact and real-time monitoring, can effectively improve the accuracy and efficiency of weld defect detection, and solves the problem of insufficient accuracy of weld detection in the prior art.

[0035] Preferably, the step of irradiating the weld to be inspected with coherent light in a sequence of unloaded and dynamic environmental loads to generate a reference state hologram and a loaded state hologram sequence comprises: S11: Maintaining the weld to be inspected in a stable state, emitting coherent light to the weld to be inspected in the stable state through a pulsed laser, so that the coherent light is split into object light that reaches a camera after irradiating the weld to be inspected and reference light that directly reaches the camera through a beam splitter, and the object light and the reference light interfere with each other on a target surface of the camera to form a reference state hologram; wherein the reference state hologram is used to describe holographic information displayed by the weld to be inspected in a no-load state; S12: applying microsecond thermal pulses and low-frequency mechanical vibrations that meet the specified specifications of the environmental load to the weld to be inspected, so that the weld to be inspected is in a deformation excitation state, and emitting coherent light to the weld to be inspected in the deformation excitation state by a pulsed laser to obtain a loaded state hologram; S13: Continuously adjust the environmental load applied to the weld to be inspected based on the dynamic environmental load sequence, so that the weld to be inspected is in a deformation excitation state corresponding to each environmental load in turn, so as to obtain a loading state hologram corresponding to the environmental load, which is collectively used as a loading state hologram sequence.

[0036] Specifically, to ensure that the weld to be inspected is in a stable state without external forces or environmental loads, the weld can be placed in a fixed position and ensured to be in a state of normal temperature and pressure to avoid external interference. The weld in a stable state is the initial state that is not affected by any load. Ensuring that the weld is holographically imaged in this state is to obtain basic reference information. The hologram obtained in the stable state can truly reflect the shape and stress distribution of the weld when it is not affected by the load, and provide benchmark data for subsequent defect detection and comparison of loading states.

[0037] More specifically, a pulsed laser is used to emit coherent light, which is then split into two parts by a beam splitter. The object light is irradiated onto the weld to be inspected and reaches the camera through reflection or scattering, while the reference light is directly irradiated onto the target surface of the camera. The emission of coherent light is to ensure that the object light and the reference light are in the same phase, thereby forming interference fringes on the target surface of the camera and obtaining high-quality holographic information. Through the interference effect, an accurate reference state hologram can be obtained, reflecting the slight deformation or irregularity of the weld to be inspected in the unloaded state, which is crucial for subsequent comparative analysis.

[0038] More specifically, the object light and the reference light interfere on the target surface of the camera to form a reference state hologram. This hologram displays the holographic information of the weld to be inspected under no-load conditions, reflecting its geometric shape and stress distribution in its original state. The reference state hologram is the benchmark under no-load conditions and serves as the basis for comparison with subsequent loaded state holograms. By accurately acquiring the reference state hologram, the deformation of the weld under no-load conditions can be analyzed in detail, providing comparison data for subsequent analysis of changes under loaded conditions.

[0039] More specifically, microsecond thermal pulses and low-frequency mechanical vibrations that meet the specified specifications of the environmental load are applied to the weld to be inspected. At this time, the weld is in a deformation excitation state and is affected by both temperature changes and mechanical vibrations. By simulating the actual environmental loads that the weld may withstand, it is put in an excitation state, which can better reflect the deformation characteristics and stress response of the weld under dynamic loads. This step can reveal fatigue damage, cracks or other defects that may occur in the weld during actual use by simulating the loads in the actual working environment, which makes defect detection closer to practical applications.

[0040] More specifically, when the weld to be inspected is in a deformation-excited state, coherent light is emitted again through a pulsed laser to illuminate the weld and obtain a loaded hologram. In order to obtain a hologram of the weld under the loaded state, coherent light must be illuminated when the weld is subjected to environmental loads to record its deformation. By obtaining the loaded hologram, the stress distribution and deformation of the weld under the deformation-excited state can be revealed, which helps analyze the defect evolution and potential problems of the weld under loading conditions.

[0041] More specifically, according to the dynamic environmental load sequence, the environmental load applied to the weld is continuously adjusted. After each adjustment, the weld will be in a new deformation excitation state. For the weld under each environmental load condition, coherent light is emitted and the corresponding loaded state hologram is obtained. This process is repeated until a series of loaded state holograms are obtained. The weld will experience different load conditions in the actual environment. Therefore, it is necessary to adjust the environmental load to simulate the response of the weld under a variety of different environmental conditions. By obtaining multiple loaded state holograms, the performance of the weld under different loads can be comprehensively evaluated, and then its deformation resistance, defect evolution and crack propagation can be analyzed.

[0042] More specifically, the holograms obtained under each loading state are aggregated to form a hologram sequence. Each loaded state hologram corresponds to the deformation state of the weld under different environmental loads. Generating a loaded state hologram sequence can show the deformation process of the weld under different loads, providing comparative information to help further assess the damage and fatigue of the weld. The loaded state hologram sequence can clearly show the deformation characteristics of the weld under different load conditions, revealing possible fatigue damage or stress concentration areas in the weld, and providing more accurate data for defect location and quality control.

[0043] It can be understood that by generating reference state and loaded state holograms and their sequences, the deformation and defects of the weld under different loads can be analyzed in detail. The acquisition of the loaded state hologram sequence enables a comprehensive characterization of the dynamic load response of the weld, revealing potential defects or early signs of material failure. By applying real environmental loads (such as thermal pulses and mechanical vibrations), the detection of weld defects is ensured to be more in line with actual working conditions, thereby improving the practical application value of the detection results. By acquiring multiple holograms, accurate defect locations and defect characteristics can be provided, providing data support for subsequent repair and quality control.

[0044] Preferably, the step of calculating each phase difference field between the loaded state hologram sequence and the reference state hologram comprises: S21: reconstructing each loaded-state hologram and the reference-state hologram in the loaded-state hologram sequence by an angular spectrum method to obtain a loaded-state complex amplitude of each loaded-state hologram and a reference-state complex amplitude of the reference-state hologram; S22: performing complex amplitude conjugate multiplication calculation on each of the loaded state complex amplitudes and the reference state complex amplitude to extract the wrapped phase, and performing phase unwrapping processing on the wrapped phase to obtain a phase difference field between each of the loaded state complex amplitudes and the reference state complex amplitude.

[0045] Specifically, the angular spectrum method is used to reconstruct each loaded state hologram and the reference state hologram in the loaded state hologram sequence. The angular spectrum method is an optical reconstruction method based on Fourier transform, which is usually used for phase and amplitude reconstruction of holograms. The loaded state hologram and the reference state hologram are subjected to Fourier transform in frequency space, and the complex amplitude is obtained by inverse Fourier transform. The angular spectrum method can efficiently process the light wave propagation and interference characteristics in the hologram, and then extract the complex amplitude of the object (including amplitude and phase information) from the hologram. In this way, the loaded state hologram and the reference state hologram can be accurately reconstructed to obtain their complex amplitude. The angular spectrum method can reconstruct the physical information of the hologram very accurately, especially when the surface or structure of the object to be detected undergoes subtle changes, which can effectively improve the reconstruction accuracy and ensure that the obtained complex amplitude can accurately reflect the optical changes in the loaded state and the reference state.

[0046] More specifically, the loaded state complex amplitude and the reference state complex amplitude are each subjected to complex amplitude conjugate multiplication. This complex amplitude conjugate multiplication is performed to extract the loaded phase (wrapped phase). By performing conjugate multiplication on the complex amplitudes, phase information can be extracted in the complex domain, i.e., the phase difference associated with the surface or structural deformation of the object. This complex amplitude conjugate multiplication simplifies complex optical information into a field with a clear phase difference, extracting the phase variation contained in the interference fringes. This operation is the basis for obtaining the phase difference field, which clearly demonstrates the phase difference between the loaded and reference states.

[0047] More specifically, the wrapped phase obtained from the conjugate multiplication calculation is subjected to phase unwrapping. Due to the phase wrapping property, the obtained phase value usually jumps within a certain range and needs to be restored to a continuous phase difference field through a phase unwrapping algorithm (for example, a phase recovery algorithm based on neighborhood search or image processing). The phase obtained from the hologram is usually a wrapped phase within a limited range. In order to obtain an accurate phase difference value, it is necessary to restore a continuous phase difference field through unwrapping technology. This process can eliminate the discontinuity caused by phase wrapping, so that the phase difference field can truly reflect the accuracy of the object deformation. The phase difference field after phase unwrapping can clearly reveal the phase difference between the loaded state and the reference state, ensuring that the details of the weld deformation can be accurately quantified in subsequent analysis. The unwrapped phase information will be more accurate and continuous, facilitating subsequent deformation analysis, defect detection, and stress analysis.

[0048] More specifically, after completing the above processing, the phase difference field between the loaded state complex amplitude and the reference state complex amplitude is obtained. This phase difference field can be analyzed by the phase difference between each loaded state hologram and the reference state hologram, and finally the phase difference field of the entire loaded state hologram sequence and the reference state hologram is obtained. The phase difference field can intuitively show the weld deformation caused by external environmental loading. In defect detection, the change in the phase difference field often reflects the presence of tiny cracks, fatigue damage or other defects in the weld. By obtaining the phase difference field between the loaded state and the reference state, the deformation or defect response of the weld under the loading environment can be quantified, which provides an accurate basis for subsequent defect detection, material fatigue analysis and quality control.

[0049] It can be understood that through the angular spectrum method and complex amplitude conjugate multiplication calculation, the phase difference between the loaded state and the reference state can be extracted with high precision, revealing the slight deformation of the object. The phase unwrapping process eliminates the jump of the phase value, so that the phase difference field remains continuous in space, providing real data for subsequent deformation analysis. The phase difference field can reveal subtle changes in the weld, especially the deformation difference under the loaded state, and can provide clear guidance for the detection and quality control of weld defects. By analyzing the phase difference under different loading conditions, the deformation of the weld under various environmental loads can be comprehensively evaluated, thereby improving the reliability and safety of the structure.

[0050] Preferably, the step of calculating the local deformation gradient of each phase difference field to mark the first abnormal area of ​​the weld to be inspected and obtaining the corresponding deformation concentration and spatial distribution pattern includes: S23: converting each phase difference field into a corresponding displacement field, and performing displacement gradient tensor decomposition on each displacement field to generate a displacement gradient tensor of each displacement field; S24: extracting a shear strain and a normal strain from the displacement gradient tensor as deformation indices to serve as local deformation gradient features of each phase difference field; S25: dispatching a corresponding deformation gradient detection standard from a prepared database according to the environmental load of the loaded state hologram corresponding to the phase difference field, so as to perform a threshold detection on the local deformation gradient feature; S26: marking an abnormal area on the loaded state hologram according to the result of the threshold detection, and performing a weighted confidence evaluation on the abnormal area fed back by each loaded state hologram to mark a first abnormal area and a corresponding weighted confidence deformation gradient feature on the reference state hologram; S27: performing spatial clustering processing on the weighted confidence deformation gradient feature to obtain the deformation concentration of the first abnormal area; S28: Performing geometric feature extraction and directional analysis on the weighted confidence deformation gradient feature to obtain a spatial distribution pattern of the first abnormal area.

[0051] Specifically, the calculated phase difference field is converted into a displacement field. The phase difference field is essentially the package information of the relative displacement between different points on the surface of an object. Through the relationship between the phase difference field and the wavelength, the displacement value of each position can be calculated. The phase difference field essentially contains the phase changes between the loaded state and the reference state, and these changes reflect the state of deformation of the object surface. Converting it into a displacement field can more intuitively represent the deformation of the object, which is convenient for subsequent stress and deformation analysis. Through the generation of the displacement field, the original phase difference information can be converted into actual physical displacement data, providing basic data for subsequent deformation analysis.

[0052] More specifically, the generated displacement field is decomposed into a displacement gradient tensor. This step converts the changes in the displacement field into gradient information. The displacement gradient tensor contains detailed information about the deformation, including changes in strain and stress. Specifically, the displacement gradient of each point is calculated and the magnitude and direction of the deformation are described in tensor form. The displacement gradient tensor is a key tool for describing the surface deformation of an object and can accurately reflect the degree of deformation at different locations. By decomposing the displacement gradient tensor, more detailed strain and stress information can be obtained, which helps to identify local deformation and abnormal areas. By decomposing the displacement gradient tensor, the directionality and intensity of the local deformation can be accurately obtained, which helps to detect potential defects and deformation areas in the weld.

[0053] More specifically, the shear strain and normal strain are extracted from the displacement gradient tensor, which respectively reflect the strain of the object in different directions. Shear strain reflects the local shear deformation of the object's surface, while normal strain describes the deformation perpendicular to the object's surface. Shear strain and normal strain are important indicators reflecting the local deformation of the object. By extracting these indicators, the deformation state of the object's surface under loading can be more clearly described, especially for areas with concentrated stress such as welds. These strain indicators provide a quantitative standard for the subsequent identification of abnormal areas and can accurately calibrate local deformation abnormalities.

[0054] More specifically, according to the environmental load in the loaded state hologram corresponding to the phase difference field, the corresponding deformation gradient detection standard is dispatched from the prepared database. The environmental load corresponding to the loaded state hologram has an important influence on the detection of deformation, so it is necessary to dispatch the detection standard suitable for the current environmental load. Different environmental loads will affect the deformation mode of the object, so it is necessary to adjust the detection standard according to the load conditions. By dispatching the corresponding standards, it can be ensured that the deformation gradient detection conforms to the deformation conditions under the actual load. By dispatching the deformation gradient detection standard according to the different environmental loads, the accuracy of abnormal area detection can be improved to make it conform to the actual load conditions.

[0055] More specifically, threshold detection is performed on the extracted local deformation gradient features. By setting a threshold standard, it is detected whether the local deformation gradient exceeds the threshold, and then the abnormal area is determined. Through threshold detection, areas where the deformation degree exceeds the normal range can be accurately distinguished, and areas where defects or abnormal deformation may exist can be marked. Threshold detection helps to automatically screen out abnormal areas with obvious deformation, avoid subjective errors in manual analysis, and improve detection efficiency and accuracy.

[0056] More specifically, a weighted confidence assessment is performed on the detected abnormal area. Based on the abnormal area data fed back by multiple loaded holograms, its confidence is evaluated and weighted processing is performed to determine the most likely abnormal area. Through weighted confidence assessment, the feedback information of different holograms can be integrated to avoid the error of a single hologram affecting the result. This method can improve the reliability of abnormal area marking. Weighted confidence assessment can enhance the stability of abnormal area detection, reduce the false alarm rate, and ensure that the abnormal area finally detected has a high confidence level.

[0057] More specifically, the weighted confidence deformation gradient features are spatially clustered. Through clustering algorithms (such as K-means or DBSCAN), the weighted confidence deformation gradient features are divided into several groups to identify areas with concentrated deformation. Spatial clustering can classify abnormal areas according to the characteristics of the deformation gradient, thereby accurately identifying high deformation areas in the weld, facilitating subsequent analysis and repair. Through spatial clustering processing, areas with high deformation concentration can be effectively extracted, and possible defects and potential risk areas in the weld can be located, facilitating subsequent repair and strengthening.

[0058] More specifically, geometric feature extraction and directional analysis are performed on the weighted confidence deformation gradient features to analyze the geometric shape and distribution direction of the deformation anomaly area. This can reveal the deformation pattern and directional characteristics of the anomaly area. Geometric and directional analysis helps understand the specific deformation pattern. For example, local deformation exhibits obvious linear, circular, or other specific geometric shapes, which helps further identify the anomaly type. Through geometric and directional analysis, the type, distribution, and direction of the deformation anomaly area can be identified, further enhancing the visualization and diagnosis capabilities of the anomaly area.

[0059] More specifically, through the above steps, the deformation concentration of the first abnormal area and its spatial distribution pattern are finally obtained. The deformation concentration reflects the severity of the abnormal area, while the spatial distribution pattern reveals the deformation direction and geometric characteristics of the abnormal area. The deformation concentration and spatial distribution pattern are key indicators for analyzing abnormal deformation and defects in welds. By quantifying this information, we can better understand the distribution of defects and carry out effective repairs. The extraction of deformation concentration and spatial distribution pattern provides precise guidance for defect location and repair, which is helpful for weld quality control and ensuring the safety of the structure.

[0060] It can be understood that through the calculation of local deformation gradients and threshold detection, areas with abnormal deformation can be accurately located, providing clear marks for subsequent defect detection. Weighted confidence assessment improves the accuracy and stability of detection and reduces false alarms by integrating multiple feedback information. Spatial clustering and geometric analysis can clearly calibrate areas where deformation is concentrated and their distribution patterns, thereby providing a more comprehensive deformation analysis. Geometric features and directional analysis can further reveal the specific patterns of deformation, which is helpful for accurate identification and repair of weld defects in complex environments.

[0061] Preferably, the steps of calculating the dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram by joint analysis in the space-frequency domain, fusing the abnormal features of each dominant spatial frequency, marking the second abnormal area of ​​the weld to be inspected, and obtaining the corresponding amplitude distribution and spatial continuity parameters include: S31: performing a local window Fourier transform on each loaded state hologram and the reference state hologram in the loaded state hologram sequence to obtain a dominant spatial frequency of each loaded state hologram and the reference state hologram; S32: Calculating the spectrum energy concentration, frequency offset length, and energy dispersion variation field based on the dominant spatial frequencies of each of the loaded state holograms and the reference state hologram to obtain abnormality feedback information between each of the loaded state holograms and the reference state hologram; S33: dispatching corresponding abnormal information detection standards from a prepared database according to the environmental load corresponding to each loaded state hologram, so as to perform abnormality detection on each abnormal feedback information; S34: marking an abnormal area and an abnormal frequency feature of the abnormal area on the loaded hologram according to the abnormality detection result; S35: performing feature fusion on the abnormal region and abnormal frequency characteristics of each of the loaded state holograms, so as to mark a second abnormal region and the corresponding fused abnormal frequency characteristics on the reference state hologram; S36: performing adaptive threshold segmentation on the fused abnormal frequency feature to extract the corresponding amplitude distribution; S37: Performing a connected domain analysis on the fused abnormal frequency feature, and extracting spatial distribution parameters based on the analysis results to obtain corresponding spatial continuity parameters.

[0062] Specifically, a local window Fourier transform is performed on each loaded-state hologram and reference-state hologram in the loaded-state hologram sequence. This step calculates the dominant spatial frequency of each hologram. The local window Fourier transform divides the image into multiple small areas and performs frequency analysis on each area. This can capture the spatial frequency characteristics of the image in different areas. The spatial frequency characteristics of the hologram can reveal the deformation pattern in the image. Through Fourier transform, the changes in the image can be analyzed from the frequency domain perspective, and the dominant spatial frequency can be extracted as the key feature for subsequent analysis. Fourier transform helps convert time domain information into frequency domain information, from which the frequency distribution characteristics of each area can be obtained, providing a reliable basis for detecting abnormal areas.

[0063] More specifically, based on the calculated dominant spatial frequency, the spectrum energy concentration, frequency offset length and energy dispersion change field are further calculated. These indicators can quantify the characteristics of the frequency distribution. The spectrum energy concentration measures the concentration of the spectrum, the frequency offset length reflects the range of frequency variation, and the energy dispersion describes the distribution of energy in the frequency space. These frequency domain indicators can more intuitively reveal abnormal changes in the image in the frequency domain. They are highly sensitive to abnormal areas and can effectively capture subtle changes in the image, especially in small areas such as welds. These calculation results provide more comprehensive frequency feature data for subsequent anomaly detection, which helps to improve the accuracy of anomaly detection.

[0064] More specifically, based on the environmental load information corresponding to the loaded hologram, the corresponding abnormal information detection standards are dispatched from the preliminary database. These standards provide adjusted detection thresholds and standards based on the impact of different environmental conditions (such as temperature, pressure, etc.) on the weld. Different environmental loads will produce different deformation patterns on the weld. The detection standards must match the environmental loads to improve the accuracy of anomaly detection. By dispatching appropriate anomaly detection standards, it can be ensured that the detection process takes into account the differences in environmental conditions, thereby improving the sensitivity and accuracy to anomalies.

[0065] More specifically, anomaly detection is performed on the abnormal feedback information of each loaded hologram. Based on preset standards, abnormal areas in each hologram are identified and the abnormal frequency characteristics of these areas are marked. This anomaly detection can automatically mark potential defect areas in the weld, helping to quickly identify areas of concern. The abnormal frequency characteristics can further confirm the type of anomaly in these areas. Anomaly detection can efficiently and accurately mark abnormal areas, reduce manual intervention, and improve the automation level and accuracy of detection.

[0066] More specifically, feature fusion is performed on the abnormal regions and abnormal frequency characteristics of each loading state hologram. This step combines information from holograms in different loading states, and by comparing features at different times and environments, ultimately marks a second abnormal region on the reference state hologram. Feature fusion helps combine abnormal information under different loading states, providing a more stable and comprehensive abnormal region marking. This can avoid missing potential problems due to failure to detect anomalies in a single state. By fusing abnormal features under different loading states, more comprehensive and stable abnormal region marking can be achieved, improving the accuracy of anomaly detection.

[0067] More specifically, adaptive threshold segmentation is performed on the fused abnormal frequency features. This process sets a dynamic threshold and adaptively adjusts it according to the distribution of the features, so as to accurately extract the amplitude distribution of the abnormal area. Adaptive threshold segmentation can dynamically adjust the segmentation standard according to the actual distribution of the abnormal features, avoiding the errors that may be caused by fixed thresholds. The extraction of amplitude distribution helps to analyze the severity and distribution pattern of the abnormal area. Adaptive threshold segmentation can improve the accuracy of the segmentation results and make the extraction of amplitude distribution more precise, which is conducive to the subsequent further analysis and processing of the abnormal area.

[0068] More specifically, the extracted amplitude distribution is subjected to connected domain analysis. This process identifies connected abnormal areas and can determine the spatial distribution of the abnormal areas. Based on the results of the connected domain analysis, the spatial distribution parameters of the abnormal areas, such as shape, area, boundary, etc., are extracted, and then the spatial continuity parameters are calculated. The connected domain analysis can help identify the coherence of the deformation area, and the spatial distribution parameters can quantify the distribution characteristics of the abnormal area. The spatial continuity parameters reflect whether the deformation of the abnormal area is regular. Through connected domain analysis and spatial continuity parameter extraction, we can deeply understand the spatial structure and deformation characteristics of the abnormal area, which will help to further optimize the weld repair strategy.

[0069] It can be understood that through Fourier transform and spectral energy analysis, the frequency characteristics of the weld image can be extracted, providing sensitivity analysis of deformation and abnormal areas. Through feature fusion and abnormal frequency feature extraction, the second abnormal area can be accurately calibrated and provide a basis for subsequent repair. Adaptive threshold segmentation ensures the accurate extraction of abnormal areas and avoids errors caused by fixed thresholds. Connected domain analysis and spatial continuity parameter extraction help to more deeply understand the spatial distribution characteristics of abnormal areas and improve the comprehensive recognition ability of weld defects. Through these steps, the accuracy and efficiency of weld defect detection can be effectively improved, ensuring that the detection results can reflect the true weld status and provide reliable data support for subsequent repairs.

[0070] Preferably, the step of performing defect characterization analysis on the weld to be inspected based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, and the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameter to obtain weld defect detection information includes: S41: taking the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration, and the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution, as a first group of analysis objects, so as to perform multi-physics field coupling processing on the weld to be inspected according to the first group of analysis objects; S42: taking the first abnormal area of ​​the weld to be inspected and the corresponding spatial distribution pattern, and the second abnormal area of ​​the weld to be inspected and the corresponding spatial continuity parameter, as a second set of analysis objects, so as to perform three-dimensional defect reconstruction of the weld to be inspected based on the second set of analysis objects; S43: Through the multi-physical field coupling processing and defect three-dimensional reconstruction, the specific defect morphology of the weld to be inspected is digitally reproduced and the stress concentration factor is analyzed to obtain a digital model of the defect morphology of the weld to be inspected and a corresponding set of stress concentration factors, which are used together as weld defect detection information.

[0071] Specifically, the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration, as well as the second abnormal area and the corresponding amplitude distribution, are taken as the first group of analysis objects. Based on this group of analysis objects, multi-physics field coupling processing is performed. This step mainly combines the deformation concentration and amplitude distribution information of the weld with other physical fields (such as mechanics, thermal, electromagnetic, etc.) for linkage simulation. The deformation concentration and amplitude distribution reflect the local strain and energy distribution of the weld under loading. By coupling this information with other data from the multi-physics field, the actual working state of the weld and the possible defect characteristics can be more comprehensively evaluated. Multi-physics field coupling can reveal the overall response of the weld under different physical effects and increase the depth of analysis of defect characteristics. The multi-physics field coupling simulation combined with deformation concentration and amplitude distribution can simulate the performance of the weld under different environments and loading conditions, which helps to comprehensively analyze the potential impact of weld defects. This method can provide comprehensive feedback on weld defects, thereby improving the ability to predict the performance of the weld in actual use environments.

[0072] More specifically, the first abnormal area of ​​the weld to be inspected and the corresponding spatial distribution pattern and the second abnormal area and the corresponding spatial continuity parameters are taken together as the second group of analysis objects. Based on this group of analysis objects, three-dimensional reconstruction of the weld defects is performed. This process constructs a three-dimensional model of the weld defects by analyzing the spatial distribution pattern and spatial continuity parameters. The spatial distribution pattern and spatial continuity parameters can describe the geometric shape and spatial distribution characteristics of the defect area. Through three-dimensional reconstruction, the spatial morphology of the weld defects and their distribution in the weld can be intuitively displayed, providing an intuitive defect model for subsequent analysis. The three-dimensional reconstruction can provide a morphologically accurate defect digital model, which is convenient for further analysis of the morphology of the weld defects. This three-dimensional model can help accurately locate the defect area and facilitate the formulation of a more reasonable repair plan.

[0073] More specifically, through multi-physics field coupling processing and three-dimensional reconstruction of defects, the specific morphology of the defects in the weld to be inspected is digitally reproduced, and the stress concentration factor of the weld defect is analyzed to evaluate the stress distribution in the defect area and its impact on the overall strength of the weld. Digital reproduction can accurately reproduce the morphology of the defect, so that subsequent analysis can be directly carried out on the actual defect characteristics. The stress concentration factor is a key parameter for evaluating the stress borne by the material at the defect. It can predict the possible failure mode of the weld under working conditions and guide defect repair. Through digital reproduction, the specific form of the defect can be intuitively displayed, which is convenient for subsequent optimization design and repair measures. The stress concentration factor analysis helps to understand the impact of defects on the weld and accurately evaluate the safety and durability of the weld.

[0074] More specifically, by combining the analysis results of the previous steps, a digital model of the defect morphology and a set of stress concentration factors of the weld to be inspected are obtained. These two results are used together as weld defect detection information for subsequent decision-making and weld repair. Through the combination of the defect morphology model and the stress concentration factor, the shape and size of the defect area and its impact on the weld performance can be fully described. The provided detection information can help formulate more targeted repair plans. Through the combination of the digital model and the stress factor set, the impact of defects on the weld can be fully understood, providing accurate data support for subsequent analysis and repair work. By comprehensively and accurately characterizing the defects, a more scientific and effective decision-making basis can be provided.

[0075] It can be understood that through the combination of multi-physical field coupling and spatial reconstruction, the causes and manifestations of weld defects can be comprehensively evaluated, and their changes under various environmental factors can be revealed. The digital reproduction of defect morphology and three-dimensional reconstruction provide accurate modeling of the defect area, which is convenient for subsequent analysis and repair. Through the analysis of stress concentration factors, the impact of defect areas on weld strength can be evaluated, thereby providing strong support for the formulation of repair plans. Through the combination of digital models and stress factor sets, comprehensive and detailed weld defect detection information can be provided, providing a scientific basis for subsequent work. Through these steps, the detection and characterization of weld defects can be fully carried out, ensuring the efficiency and accuracy of the detection results, and providing reliable support for the repair and optimization of welds.

[0076] Reference Figure 2 As shown, in a second aspect, the present invention provides a weld defect detection device based on three-dimensional holographic scanning, which is used to implement a weld defect detection method based on three-dimensional holographic scanning as described in any one of the first aspects, comprising: A dynamic testing module is used to irradiate the weld to be inspected with coherent light in a sequence of unloaded and dynamic environmental loads to generate a reference state hologram and a loaded state hologram sequence; a phase analysis module, configured to calculate each phase difference field between the loaded state hologram sequence and the reference state hologram, and to calculate a local deformation gradient for each phase difference field, so as to mark a first abnormal area of ​​the weld to be inspected and obtain a corresponding deformation concentration and spatial distribution pattern; a frequency analysis module for calculating the dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram through joint analysis in the space-frequency domain, so as to fuse the abnormal features of each dominant spatial frequency, mark the second abnormal area of ​​the weld to be inspected, and obtain the corresponding amplitude distribution and spatial continuity parameters; The defect analysis module is used to perform defect characterization analysis on the weld to be inspected based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, as well as the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameters, so as to obtain weld defect detection information.

[0077] In this embodiment, for the specific implementation of each module in the above system embodiment, please refer to the above method embodiment, which will not be repeated here.

[0078] In a third aspect, the present invention provides a weld defect detection device based on three-dimensional holographic scanning, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements a weld defect detection method based on three-dimensional holographic scanning as described in any one of the first aspects.

[0079] 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 and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A weld defect detection method based on three-dimensional holographic scanning, characterized in that: include: The weld to be inspected is irradiated with coherent light in a sequence of no load and dynamic environmental load to generate a reference state hologram and a loaded state hologram sequence; Calculating each phase difference field between the loaded state hologram sequence and the reference state hologram, and calculating the local deformation gradient of each phase difference field to mark the first abnormal area of ​​the weld to be inspected and obtain the corresponding deformation concentration and spatial distribution pattern; The dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram are calculated by joint analysis in the space-frequency domain to fuse the abnormal features of each dominant spatial frequency, so as to mark the second abnormal area of ​​the weld to be inspected and obtain the corresponding amplitude distribution and spatial continuity parameters; Based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, as well as the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameters, a defect characterization analysis is performed on the weld to be inspected to obtain weld defect detection information.

2. The weld defect detection method based on three-dimensional holographic scanning according to claim 1, characterized in that: The steps of irradiating the weld to be inspected with coherent light in an unloaded state and a dynamic environmental load sequence to generate a reference state hologram and a loaded state hologram sequence include: The weld to be inspected is maintained in a stable state, and coherent light is emitted to the weld to be inspected in the stable state by a pulsed laser. The coherent light is then split by a beam splitter into object light that reaches a camera after irradiating the weld to be inspected, and reference light that directly reaches the camera. The object light and the reference light interfere with each other on a target surface of the camera to form a reference state hologram. The reference state hologram is used to describe the holographic information displayed by the weld to be inspected when no load is applied. Applying microsecond thermal pulses and low-frequency mechanical vibrations that meet the specified specifications of the environmental load to the weld to be inspected, so that the weld to be inspected is in a deformation excitation state, and emitting coherent light to the weld to be inspected in the deformation excitation state through a pulsed laser to obtain a loaded state hologram; Based on the dynamic environmental load sequence, the environmental load applied to the weld to be inspected is continuously adjusted so that the weld to be inspected is in a deformation excitation state corresponding to each environmental load in turn, so as to obtain a loaded state hologram corresponding to the environmental load, which is collectively used as a loaded state hologram sequence.

3. The weld defect detection method based on three-dimensional holographic scanning according to claim 1, characterized in that: The step of calculating each phase difference field between the loaded state hologram sequence and the reference state hologram comprises: Reconstructing each loaded-state hologram and the reference-state hologram in the loaded-state hologram sequence by an angular spectrum method to obtain a loaded-state complex amplitude of each loaded-state hologram and a reference-state complex amplitude of the reference-state hologram; Complex amplitude conjugate multiplication is performed on each of the loaded state complex amplitudes and the reference state complex amplitude to extract the wrapped phase, and phase unwrapping is performed on the wrapped phase to obtain a phase difference field between each of the loaded state complex amplitudes and the reference state complex amplitude.

4. The weld defect detection method based on three-dimensional holographic scanning according to claim 1, characterized in that: The steps of calculating the local deformation gradient of each phase difference field to mark the first abnormal area of ​​the weld to be inspected and obtain the corresponding deformation concentration and spatial distribution pattern include: Converting each phase difference field into a corresponding displacement field, and performing displacement gradient tensor decomposition on each displacement field to generate a displacement gradient tensor of each displacement field; Extracting shear strain and normal strain from the displacement gradient tensor as deformation indices to serve as local deformation gradient features of each phase difference field; According to the environmental load of the loaded state hologram corresponding to the phase difference field, the corresponding deformation gradient detection standard is dispatched from the prepared database to perform threshold detection on the local deformation gradient feature; Marking an abnormal area on the loaded state hologram according to the result of the threshold detection, and performing a weighted confidence evaluation on the abnormal area fed back by each loaded state hologram to mark a first abnormal area and a corresponding weighted confidence deformation gradient feature on the reference state hologram; Performing spatial clustering processing on the weighted confidence deformation gradient features to obtain the deformation concentration of the first abnormal area; Geometric feature extraction and directionality analysis are performed on the weighted confidence deformation gradient feature to obtain a spatial distribution pattern of the first abnormal area.

5. The weld defect detection method based on three-dimensional holographic scanning according to claim 1, characterized in that: The steps of calculating the dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram by joint analysis in the space-frequency domain, fusing the abnormal features of each dominant spatial frequency, marking the second abnormal area of ​​the weld to be inspected, and obtaining the corresponding amplitude distribution and spatial continuity parameters include: Performing a local window Fourier transform on each loaded state hologram and the reference state hologram in the loaded state hologram sequence to obtain a dominant spatial frequency of each loaded state hologram and the reference state hologram; Calculating the spectrum energy concentration, frequency offset length, and energy dispersion variation field based on the dominant spatial frequencies of each of the loaded state holograms and the reference state hologram to obtain abnormality feedback information between each of the loaded state holograms and the reference state hologram; According to the environmental load corresponding to each of the loaded state holograms, the corresponding abnormal information detection standard is dispatched from the prepared database to perform abnormality detection on each of the abnormal feedback information; Marking an abnormal area and an abnormal frequency feature of the abnormal area on the loaded state hologram according to the result of the abnormality detection; Performing feature fusion on the abnormal region and abnormal frequency characteristics of each of the loaded state holograms to mark a second abnormal region and the corresponding fused abnormal frequency characteristics on the reference state hologram; Performing adaptive threshold segmentation on the fused abnormal frequency features to extract corresponding amplitude distribution; A connected domain analysis is performed on the fused abnormal frequency feature, and spatial distribution parameters are extracted based on the analysis results to obtain corresponding spatial continuity parameters.

6. The weld defect detection method based on three-dimensional holographic scanning according to claim 1, characterized in that: The step of performing defect characterization analysis on the weld to be inspected based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, and the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameter to obtain weld defect detection information includes: The first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration, and the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution are taken as a first group of analysis objects, so as to perform multi-physics field coupling processing on the weld to be inspected according to the first group of analysis objects; The first abnormal area of ​​the weld to be inspected and the corresponding spatial distribution pattern, and the second abnormal area of ​​the weld to be inspected and the corresponding spatial continuity parameter are used together as a second group of analysis objects, so as to perform three-dimensional defect reconstruction of the weld to be inspected based on the second group of analysis objects; Through the multi-physical field coupling processing and defect three-dimensional reconstruction, the specific defect morphology of the weld to be inspected is digitally reproduced and the stress concentration factor is analyzed to obtain a digital model of the defect morphology of the weld to be inspected and a corresponding set of stress concentration factors, which are used together as weld defect detection information.

7. A weld defect detection device based on three-dimensional holographic scanning, characterized in that: A method for detecting weld defects based on three-dimensional holographic scanning, for implementing any one of claims 1 to 6, comprising: A dynamic testing module is used to irradiate the weld to be inspected with coherent light in a sequence of unloaded and dynamic environmental loads to generate a reference state hologram and a loaded state hologram sequence; a phase analysis module, configured to calculate each phase difference field between the loaded state hologram sequence and the reference state hologram, and to calculate a local deformation gradient for each phase difference field, so as to mark a first abnormal area of ​​the weld to be inspected and obtain a corresponding deformation concentration and spatial distribution pattern; a frequency analysis module for calculating the dominant spatial frequencies of the loaded state hologram sequence and the reference state hologram through joint analysis in the space-frequency domain, so as to fuse the abnormal features of each dominant spatial frequency, mark the second abnormal area of ​​the weld to be inspected, and obtain the corresponding amplitude distribution and spatial continuity parameters; The defect analysis module is used to perform defect characterization analysis on the weld to be inspected based on the first abnormal area of ​​the weld to be inspected and the corresponding deformation concentration and spatial distribution pattern, as well as the second abnormal area of ​​the weld to be inspected and the corresponding amplitude distribution and spatial continuity parameters, so as to obtain weld defect detection information.

8. A weld defect detection device based on three-dimensional holographic scanning, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the method for detecting weld defects based on three-dimensional holographic scanning as described in any one of claims 1 to 6 is implemented.

Images