Thin-walled nickel alloy pipe inner wall hydraulic test detection system and method

Abstract

The present application relates to the technical field of nickel alloy pipe water pressure detection, and discloses a thin-walled nickel alloy pipe inner wall hydraulic test detection system and method. The method comprises establishing a closed water pressure environment in the pipe and applying a preset pressure gradient, collecting pipe wall deformation response data through a distributed sensing array; generating a composite feature matrix containing axial strain, circumferential stress and radial displacement features through multi-dimensional feature extraction, inputting a material performance degradation evaluation model, simulating microstructure changes to output a potential defect distribution map. The water pressure gradient and pressure holding time are dynamically adjusted according to the map, a second hydraulic test is performed, and micro-deformation sequence images are captured through high-frequency imaging; the deviation amount is identified by comparing the images with the model prediction results, the model parameters are iteratively updated to optimize the detection threshold; the test is completed using the optimized threshold, and the qualified pipe section marker and defect pipe section positioning coordinates are output, thereby improving the comprehensiveness and accuracy of thin-walled nickel alloy pipe inner wall defect detection.

Description

Technical Field

[0001] This invention relates to the field of hydrostatic testing technology for nickel alloy pipes, specifically to a hydrostatic testing system and method for the inner wall of thin-walled nickel alloy pipes. Background Technology

[0002] Thin-walled nickel alloy tubes, with their excellent high-temperature resistance, corrosion resistance, and high strength, are widely used in critical fields such as fuel delivery pipelines for aerospace engines, cooling pipes for nuclear power reactors, and high-pressure media transmission pipelines for chemical industries. The quality of the inner wall of these tubes directly affects the operational safety of the entire system. Due to their thin wall thickness (typically within the range of 0.5-3mm), potential defects such as microcracks, intergranular corrosion, and localized thinning can easily expand rapidly under long-term high-temperature and high-pressure operating conditions, leading to media leakage or even pipe rupture, causing serious safety accidents and economic losses. Therefore, accurate detection of defects in the inner wall of thin-walled nickel alloy tubes is an important prerequisite for ensuring the reliable operation of related equipment.

[0003] Currently, hydrostatic testing is a commonly used method for testing the sealing and structural integrity of the inner walls of thin-walled nickel alloy tubes. However, existing technologies have many limitations. Traditional hydrostatic testing often uses the presence or absence of macroscopic leakage as the criterion, which can only identify through-hole defects or surface defects with large pore sizes. It cannot capture potential microscopic defects inside the material that have not yet formed through-hole channels, such as microcracks with a depth of less than 0.1 mm or local grain boundary weakening regions. These potential defects are likely to become sources of failure during subsequent service, leading to a significant risk of missed detection by traditional testing methods.

[0004] In the sensor data acquisition stage, existing water pressure tests mostly use single-point strain sensors or pressure sensors, which can only acquire local data at the sensor installation location and cannot achieve comprehensive monitoring of the overall deformation state of the pipe wall. The deformation of thin-walled nickel alloy pipes under water pressure exhibits obvious multidirectional and distributed characteristics. The changes in axial strain, circumferential stress, and radial displacement are interrelated. Single-point data is insufficient to reflect the deformation differences in different areas of the pipe wall, easily leading to misjudgment or omission of defect locations. For example, when a local microcrack exists in a certain area of ​​the pipe wall, the circumferential stress change in that area may be significantly different from other areas. However, if the single-point sensor does not cover this area, it cannot capture this characteristic difference, thus missing the opportunity to identify defects.

[0005] The prior art lacks quantitative analysis capability of material performance degradation state, and detection parameters are mostly dependent on empirical setting, such as fixed pressure gradient and pressure holding time, which cannot dynamically adjust test parameters according to the actual deformation characteristics of the pipe wall. When the pipe wall material has performance fluctuations due to batch differences or previous processing, the water pressure test with fixed parameters may cause excessive deformation of the pipe wall due to excessive pressure, or fail to stimulate the deformation response of potential defects due to insufficient pressure, affecting the accuracy and reliability of the test results. At the same time, the existing detection method does not establish a closed-loop mechanism of "detection result-model optimization", and once the detection threshold is set, it is fixed for a long time, which cannot adapt to the detection needs of different service states and different defect types. For example, for the pipe that has intergranular corrosion after long-term service, the fixed threshold is easy to misjudge the slight corrosion area as qualified, or overjudge the normal deformation area as a defect, resulting in poor detection adaptability.

[0006] The existing water pressure test is mostly a single test process, and lacks a secondary verification link of the test results. The data obtained by only one water pressure loading is difficult to exclude accidental factors interference, such as local deformation anomaly caused by pipe wall surface stains, or error data caused by temporary sensor failure. These interference factors are easy to reduce the reliability of the test results, and cannot provide reliable basis for subsequent pipe segment screening and use. SUMMARY

[0007] The purpose of the present application is to provide a thin-walled nickel alloy pipe inner wall water pressure test detection system and method to solve the problems raised in the background art.

[0008] To achieve the above purpose, the present application provides a thin-walled nickel alloy pipe inner wall water pressure test detection method, which comprises:

[0009] A closed water pressure environment is established inside the thin-walled nickel alloy pipe, and a preset pressure gradient is applied. The pipe wall deformation response data is collected by a distributed sensor array;

[0010] The pipe wall deformation response data is subjected to multi-dimensional feature extraction to generate a composite feature matrix containing axial strain features, circumferential stress features and radial displacement features;

[0011] A thin-walled nickel alloy pipe material performance degradation evaluation model is constructed, the composite feature matrix is input into the model for material microstructure change simulation, and a potential defect distribution map is output;

[0012] According to the potential defect distribution map, a water pressure loading strategy adjustment instruction is generated to dynamically adjust the pressure gradient and pressure holding time parameters in the water pressure environment;

[0013] Based on the adjusted water pressure parameters, a secondary water pressure test is performed, and the pipe wall micro-deformation sequence images are captured by a high-frequency imaging unit;

[0014] The micro-deformation sequence image is compared with the prediction result of the material performance degradation evaluation model, and the deviation amount of the actual defect area and the predicted defect area is identified;

[0015] The material performance degradation evaluation model is iteratively updated according to the deviation amount, and an optimized defect detection threshold range is generated;

[0016] The optimized defect detection threshold range is used for final water pressure test verification of the thin-walled nickel alloy pipe, and a qualified pipe segment mark and a defect pipe segment positioning coordinate are output.

[0017] Preferably, the pipe wall deformation response data collected by the distributed sensing array comprises:

[0018] Three groups of optical fiber strain sensors are arranged axially on the thin-walled nickel alloy pipe, each group containing twelve sensing nodes distributed at equal intervals;

[0019] All sensing nodes are synchronously activated and the initial water pressure loading stage pipe wall basic deformation data is recorded, and the sampling frequency is not less than 1 kHz;

[0020] The dynamic deformation response waveform is collected in the water pressure gradient rising stage, and the stable holding time of each pressure gradient step is recorded;

[0021] The dynamic deformation response waveform is processed in time domain to eliminate the baseline drift caused by the difference in sensor installation position;

[0022] The processed dynamic deformation response waveform is spatio-temporally associated with the corresponding pressure gradient value to generate a pipe wall deformation response data set with pressure-deformation mapping relationship.

[0023] Preferably, the multi-dimensional feature extraction of the pipe wall deformation response data comprises:

[0024] The axial strain waveform, circumferential stress waveform and radial displacement waveform are separated from the pipe wall deformation response data set as three independent data channels;

[0025] The time domain statistical features, frequency energy features and nonlinear dynamic features are extracted for each data channel respectively;

[0026] The strain-stress correlation feature vector is constructed using the time domain statistical features, including peak factor, waveform index and pulse index;

[0027] The resonance frequency band feature vector is constructed using the frequency energy features, including dominant frequency component, harmonic energy ratio and frequency band power spectral density;

[0028] The nonlinear dynamic features are converted into phase space reconstruction parameters to generate a dynamic feature vector representing the nonlinear response of the material;

[0029] The strain-stress correlation feature vector, the resonance frequency band feature vector and the dynamic feature vector are fused according to a time window to form a composite feature matrix.

[0030] Preferably, the thin-walled nickel alloy pipe material performance degradation evaluation model comprises:

[0031] A multi-scale material response simulation framework comprising a lattice distortion module, a dislocation motion module and a micro-crack propagation module is established.

[0032] The strain-stress correlation feature vector in the composite feature matrix is input into the lattice distortion module to calculate the theoretical lattice strain energy density distribution.

[0033] The resonance frequency band feature vector is input into the dislocation motion module to simulate the local energy dissipation mode caused by the dislocation pinning effect.

[0034] The dynamic feature vector drives the micro-crack propagation module to run to predict the change trajectory of the crack tip stress field intensity factor.

[0035] The output results of the three modules are integrated to generate a potential defect distribution map reflecting the micro-damage evolution process of the material.

[0036] In the potential defect distribution map, the area with a stress concentration coefficient exceeding a critical value is marked as a high-risk defect area.

[0037] Preferably, the water pressure loading strategy adjustment instruction generated according to the potential defect distribution map comprises:

[0038] The spatial distribution characteristics and stress concentration degree of the high-risk defect area in the potential defect distribution map are analyzed.

[0039] A stepped pressure increase instruction is generated for the dense defect area, and a plurality of intermediate pressure verification nodes are set.

[0040] A pulse loading instruction is generated for the isolated defect area, and a short-time high-pressure impact mode is used to stimulate the defect response.

[0041] The pressure holding time parameter is adjusted according to the spatial density distribution of the defect area, and the pressure holding time is prolonged in the high-density area.

[0042] The stepped pressure increase instruction, the pulse loading instruction and the pressure holding time adjustment parameter are integrated into a water pressure loading strategy adjustment instruction set.

[0043] Preferably, the pipe wall micro-deformation sequence image captured by the high-frequency imaging unit comprises:

[0044] A high-speed digital image correlation measurement system is deployed on the outer surface of the thin-walled nickel alloy pipe, and the sampling resolution reaches the micron level.

[0045] According to the water pressure loading strategy adjustment instruction set, a pressure node is set, and three-dimensional digital image acquisition is triggered synchronously;

[0046] Twenty full-field strain distribution images are continuously captured for each pressure node, and the frame interval is not greater than 5 ms;

[0047] Sub-pixel displacement tracking algorithm is used to process the sequence images, and the three-dimensional displacement field time-varying process of the pipe wall surface is reconstructed;

[0048] Three feature image layers of maximum principal strain distribution, minimum principal strain distribution and shear strain distribution are extracted from the three-dimensional displacement field;

[0049] The feature image layers are arranged in order of pressure gradient, and a micro deformation sequence image database with spatial and temporal continuity is constructed.

[0050] Preferably, the difference comparison between the micro deformation sequence image and the prediction result of the material performance degradation evaluation model includes:

[0051] The maximum principal strain distribution image is extracted from the micro deformation sequence image database as an actual observation data set;

[0052] The theoretical strain distribution graph is derived from the potential defect distribution graph output by the material performance degradation evaluation model;

[0053] The pixel-level correspondence between the actual observation data set and the theoretical strain distribution graph is established, and the local strain difference field is calculated;

[0054] The local strain difference field is subjected to regional clustering analysis, and the abnormal response area with a difference amplitude exceeding the standard deviation is identified;

[0055] The area ratio, geometric shape factor and spatial distribution dispersion of the abnormal response area are calculated as three key difference indicators;

[0056] According to the key difference indicators, a model prediction deviation quantitative report is generated, and the defect identification parameters that need to be corrected are marked.

[0057] Preferably, the parameter iterative updating of the material performance degradation evaluation model according to the deviation amount includes:

[0058] The defect identification parameter correction requirement marked in the model prediction deviation quantitative report is analyzed;

[0059] The elastic constant tensor parameters of the lattice distortion module are adjusted to match the strain gradient distribution of the actual observation data set;

[0060] The critical shear stress threshold of the dislocation motion module is corrected, and the geometric shape factor of the abnormal response area is recalibrated;

[0061] The crack initiation criterion of the micro crack propagation module is optimized, and the stress intensity factor calculation formula is updated according to the spatial distribution dispersion;

[0062] Through the cooperative optimization of the three modules, the correlation coefficient between the theoretical strain distribution diagram output by the model and the actual observation data set is improved to above the predetermined value;

[0063] The optimized parameters are configured and stored, and a new defect detection threshold range is generated and applied to the final hydrostatic test verification stage.

[0064] Preferably, the final hydrostatic test verification of the thin-walled nickel alloy pipe using the optimized defect detection threshold range comprises:

[0065] The pressure test is carried out under the condition that the rated working pressure of the thin-walled nickel alloy pipe is one and a half times, and the duration is not less than ten minutes;

[0066] The pipe wall deformation response data is monitored in real time and compared with the optimized defect detection threshold range point by point;

[0067] When the deformation response of a certain area continuously exceeds the threshold range, the area coordinates are immediately marked and the overrun parameters are recorded;

[0068] After the test is completed, the spatial position information and parameter deviation degree of all overrun areas are summarized;

[0069] According to the distribution characteristics of the overrun areas, the thin-walled nickel alloy pipe is divided into three grades of qualified pipe section, repairable pipe section and scrap pipe section;

[0070] The final detection report containing the pipe section grade mark and the defect coordinate list is output.

[0071] Preferably, the present application also includes a thin-walled nickel alloy pipe inner wall hydrostatic test detection system, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, and the processor implements the steps of the above-mentioned thin-walled nickel alloy pipe inner wall hydrostatic test detection method when executing the computer program.

[0072] Compared with the prior art, the present application has the following beneficial effects:

[0073] By establishing a closed water pressure environment inside the thin-walled nickel alloy pipe and combining it with the distributed sensing array to collect pipe wall deformation response data, the limitations of traditional single-point sensing are broken. The distributed sensing array can achieve full coverage monitoring of the overall area of the pipe wall, capturing the deformation differences at different positions under water pressure. Whether it is the subtle changes in axial strain, uneven distribution of circumferential stress, or local fluctuations in radial displacement, they can all be fully collected, avoiding the potential omission of defects caused by insufficient sensing coverage. It provides more complete and representative basic data for subsequent defect analysis, and upgrades the detection process from "local monitoring" to "overall perception".

[0074] The multi-dimensional feature extraction link generates a composite feature matrix containing axial strain features, circumferential stress features, and radial displacement features. Compared with traditional single feature analysis, it can more comprehensively depict the stress and deformation state of the pipe wall material. The deformation of the thin-walled nickel alloy pipe under water pressure is a comprehensive manifestation of multi-directional mechanical response. Single feature can only reflect a certain aspect of the state, while the composite feature matrix integrates the three key features to construct a feature model that is closer to the actual stress condition of the material. For example, the axial strain feature can reflect the overall stretching state of the pipe section, the circumferential stress feature can highlight the local structure weak area, and the radial displacement feature can assist in judging whether there are protrusions or concave defects on the inner wall. The combination of the three makes the feature information more abundant, providing a more reliable basis for subsequent material performance analysis.

[0075] The construction and application of the material performance degradation evaluation model realize the leap from "macroscopic deformation observation" to "microstructure simulation". This model can infer the changes in the internal microstructure of the material based on the composite feature matrix, rather than just observing the surface deformation. It can effectively identify potential problems inside the material that have not yet manifested as macroscopic defects, such as micro-crack initiation and local performance weakening caused by grain boundary corrosion. By outputting the potential defect distribution map, it can clearly present the approximate location and distribution range of the defects, providing a clear direction for subsequent water pressure parameter adjustment, avoiding the "blind test" caused by the lack of micro-analysis in traditional detection, and making the detection process more targeted.

[0076] According to the potential defect distribution map, the water pressure loading strategy can be dynamically adjusted to break free from the constraints of traditional fixed parameters and achieve "on-demand adjustment". For the potential defect areas shown in the map, the deformation response of this area can be enhanced by adjusting the pressure gradient, or the deformation features of potential defects can be more fully revealed by prolonging the pressure holding time, improving the defect capture ability of the second test. For areas without obvious defects, the pressure holding time can be appropriately optimized to avoid unnecessary test time, ensuring detection effectiveness while improving test efficiency. This dynamic adjustment mechanism makes the water pressure test parameters match the actual defect state of the pipe wall, rather than mechanically applying fixed standards, significantly improving the adaptability and effectiveness of the test.

[0077] The combination of the secondary hydraulic test and the high-frequency imaging unit provides important verification and supplement for the detection results. The high-frequency imaging can capture sequence images of the microscopic deformation of the pipe wall, intuitively present the dynamic change process of the pipe wall under the adjusted water pressure parameters, verify whether the defect area predicted by the model in the first test is real, and find the small defects ignored in the first test due to the unobvious deformation. The continuity of the sequence images can also reflect the deformation trend of the defects at different water pressure stages, such as whether the micro-cracks are expanded with the increase of the pressure, to provide more basis for judging the severity of the defects, avoid misjudgment due to the contingency of the single test data, and further improve the reliability of the detection results.

[0078] Through the difference comparison between the microscopic deformation sequence images and the model prediction results, the deviation amount between the actual defect area and the predicted defect area can be accurately identified. This deviation information is not simply error data, but provides a key basis for model iteration and update. Different batches and different service states of the thin-walled nickel alloy pipe may exhibit different mechanical response characteristics due to material composition fluctuations, processing technology differences or previous service damage, and a fixed model cannot adapt to all situations. The deviation comparison provides a clear direction for model parameter adjustment, so that the model can gradually fit the actual detection scene and improve the adaptability to different types of thin-walled nickel alloy pipes.

[0079] The iterative update of the model parameters and the optimization of the detection threshold make the detection system have the ability of continuous improvement. With the accumulation of detection data, the model can continuously correct the parameters to optimize the simulation accuracy of the microstructure changes of the material, and the corresponding defect detection threshold can also more accurately match the actual defect characteristics, avoiding misjudgment or omission due to fixed threshold. For example, for a batch of thin-walled nickel alloy pipes with slight intergranular corrosion, the optimized threshold can more accurately distinguish between normal corrosion and excessive defects, so that the detection results are more in line with the actual use requirements. The optimized threshold can be directly applied in the subsequent detection process without the need for re-parameter adjustment, significantly improving the detection efficiency and consistency.

[0080] The final hydraulic test verification output of the qualified pipe segment mark and the defect pipe segment positioning coordinates provides a clear guide for subsequent pipe segment processing. The qualified pipe segment mark can be directly used to screen products that meet the requirements, and the defect pipe segment positioning coordinates can accurately point out the location of the defects, facilitating subsequent targeted repair or rejection, avoiding the overall pipe segment scrap due to vague defect positioning, and reducing resource waste. At the same time, the positioning coordinates also provide a position reference for subsequent analysis of the causes of the defects, such as the frequent occurrence of defects in a certain area, which can assist in tracing the problems in the processing technology or raw materials, providing a direction for production improvement, and realizing the linkage of detection and production optimization. BRIEF DESCRIPTION OF DRAWINGS

[0081] Figure 1The working principle diagram of the thin-walled nickel alloy pipe inner wall hydrostatic test detection method of the present application;

[0082] Figure 2 The flow chart for the distributed sensing array to collect the pipe wall deformation response data;

[0083] Figure 3 The flow chart for the high-frequency imaging unit to capture the microscopic deformation sequence images. DETAILED DESCRIPTION

[0084] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0085] Please refer to Figure 1 The present application provides a thin-walled nickel alloy pipe inner wall hydrostatic test detection method, which comprises: establishing a closed water pressure environment inside the thin-walled nickel alloy pipe and applying a preset pressure gradient, and collecting the deformation response data of the pipe wall under the action of the pressure by using a distributed sensing array. The collected original data is subjected to a multi-dimensional feature extraction process to generate a composite feature matrix containing axial strain features, circumferential stress features and radial displacement features. The composite feature matrix is input into a pre-constructed thin-walled nickel alloy pipe material performance degradation evaluation model, the model outputs a potential defect distribution map by simulating the changes of the material microscopic structure. The system automatically generates water pressure loading strategy adjustment instructions according to the map, dynamically adjusts the pressure gradient and pressure holding time parameters in the subsequent water pressure environment. Based on the adjusted parameters, a second hydrostatic test is performed, and a high-frequency imaging unit is enabled to capture the microscopic deformation sequence images of the pipe wall. The captured actual image data and the prediction results of the material performance degradation evaluation model are finely compared to identify the deviation between them. The internal parameters of the material performance degradation evaluation model are iteratively updated using the deviation to generate a set of optimized and more accurate defect detection threshold ranges. Finally, the optimized threshold ranges are used to perform a final hydrostatic test on the thin-walled nickel alloy pipe to output the marking information of the qualified pipe section and the specific positioning coordinates of the defective pipe section, thus completing the entire detection process.

[0086] Embodiment 1: refer to Figure 2, The implementation of the thin-walled nickel alloy tube inner wall hydraulic test detection method relies on high-precision data acquisition and in-depth feature analysis. The deployment of distributed sensor arrays forms the basis of physical information acquisition. In the axial direction of the thin-walled nickel alloy tube, three groups of optical fiber strain sensor groups must be uniformly arranged. Each group of optical fiber strain sensor groups integrates twelve independent sensing nodes distributed in a strictly equidistant manner. This arrangement aims to completely cover the axial strain field of the thin-walled nickel alloy tube and capture the asymmetry of the circumferential deformation. All thirty-six sensing nodes are activated synchronously with millisecond-level precision through a central synchronous controller. After synchronous activation, the initial water pressure loading stage of the thin-walled nickel alloy tube wall begins to record the basic deformation data. The sampling frequency of data acquisition is set to no less than 1 kHz, which can effectively capture the transient response that may occur during the establishment of water pressure. After entering the water pressure gradient rising stage, the system continuously collects the dynamic deformation response waveform of the thin-walled nickel alloy tube wall. The dynamic deformation response waveform reflects the elastic and plastic deformation behavior of the material under different pressure steps. At the same time, the high-precision clock inside the system records the stable holding time after each pressure gradient step is reached. The holding time data is crucial for analyzing the creep relaxation characteristics of the material.

[0087] The original collected dynamic deformation response waveform contains systematic errors caused by installation process differences. Time domain normalization processing of the dynamic deformation response waveform is a key step to eliminate baseline drift. The normalization algorithm takes the reading of each sensing node under the initial zero pressure state as the reference point, and adjusts the signal amplitude during the entire pressure loading process to a unified reference system. The processed dynamic deformation response waveform is spatio-temporally correlated with the corresponding pressure gradient value. The encoding process labels each data point with an accurate timestamp and pressure value tag, generating a clear structure and explicit pressure-deformation mapping relationship of the tube wall deformation response dataset. The tube wall deformation response dataset serves as the original data pool for subsequent feature extraction. The multi-dimensional feature extraction process separates three independent data channels from the tube wall deformation response dataset. The three data channels correspond to the axial strain waveform, circumferential stress waveform, and radial displacement waveform, respectively. The separation algorithm is based on the physical location and measurement direction characteristics of the sensor nodes. Time domain statistical features, frequency energy features, and nonlinear dynamic features are extracted in parallel for each independent data channel. Time domain statistical feature extraction calculates the peak value, mean value, variance, skewness, kurtosis, and zero-crossing rate of the waveform. These parameters describe the macroscopic statistical characteristics of the deformation signal. Frequency energy feature extraction converts the signal to the frequency domain through fast Fourier transform, analyzes the dominant frequency components, calculates the energy proportion of each harmonic, and computes the power spectral density of different frequency bands. Frequency domain features reveal the resonance characteristics and energy distribution of the material. Nonlinear dynamic feature extraction uses phase space reconstruction technology to calculate the maximum Lyapunov exponent, correlation dimension, and Kolmogorov entropy. These parameters quantify the chaos and nonlinearity of the material deformation process.

[0088] The extracted time-domain statistical features are used to construct a strain-stress correlation feature vector, which includes three core parameters, i.e., a peak factor, a waveform index, and a pulse index. The peak factor reflects the impact characteristics of the waveform, the waveform index describes the similarity between the waveform and a sine wave, and the pulse index represents the pulse intensity of the signal. The frequency-domain energy features are used to construct a resonance frequency band feature vector, which includes a dominant frequency component, a harmonic energy ratio, and a frequency band power spectral density. The dominant frequency component indicates the main vibration mode of the material, the harmonic energy ratio reflects the degree of nonlinear distortion, and the frequency band power spectral density describes the distribution of energy in the frequency domain. The nonlinear dynamics features are converted into phase space reconstruction parameters through calculation, generating a dynamics feature vector representing the nonlinear response of the material. The dynamics feature vector includes the geometric topological characteristics of the system in the phase space. The fusion process integrates the strain-stress correlation feature vector, the resonance frequency band feature vector, and the dynamics feature vector according to a pre-set time window. The sliding step of the time window matches the sampling rate of the data acquisition, ensuring the continuity of the space-time information. After standardization, the three feature vectors are spliced to form a high-dimensional composite feature matrix. The rows of the composite feature matrix correspond to time points, and the columns correspond to different feature dimensions. The composite feature matrix constitutes a comprehensive digital portrait of the material microstructure state, providing multi-angle and multi-scale input data for the subsequent performance degradation evaluation model.

[0089] Example 2: Construction of a thin-walled nickel alloy pipe material performance degradation evaluation model The construction of a thin-walled nickel alloy pipe material performance degradation evaluation model begins with the establishment of a multi-scale material response simulation framework. The multi-scale material response simulation framework logically integrates three calculation units, i.e., a lattice distortion module, a dislocation motion module, and a micro-crack propagation module, which correspond to different physical mechanisms of material microstructure damage. The strain-stress correlation feature vector included in the composite feature matrix is imported into the lattice distortion module as the main input data. The lattice distortion module maps the macroscopic strain field to the grain scale based on the theories of continuum mechanics and crystal plasticity, calculates the distribution of theoretical lattice strain energy density in the thin-walled nickel alloy pipe space, and reflects the spatial inhomogeneity of elastic deformation energy storage. The resonance frequency band feature vector is input into the dislocation motion module in parallel. The dislocation motion module uses a discrete dislocation dynamics model to simulate the slip, proliferation, and pinning effect of dislocation lines under cyclic stress, which is quantified by analyzing the attenuation characteristics of the vibration spectrum. The dynamics feature vector drives the micro-crack propagation module to run. The micro-crack propagation module is based on the principles of fracture mechanics and uses phase space reconstruction parameters to infer the initiation position of micro-cracks. It also uses the extended finite element method to predict the trajectory of the stress field intensity factor at the crack tip with the loading history, which is a key parameter for determining whether the crack enters the unstable expansion stage.

[0090] The internal calculation of the lattice distortion module relies on a set of micro-constitutive relations including elastic constant tensor, lattice orientation distribution function and grain boundary energy parameters. The elastic constant tensor defines the anisotropic elastic properties of single crystal materials, the lattice orientation distribution function describes the texture characteristics of polycrystalline materials, and the grain boundary energy parameters affect the stress concentration behavior at the grain boundaries. The simulation basis of the dislocation motion module is the preset dislocation density evolution equation, dislocation interaction rule and thermal activation energy threshold. The dislocation density evolution equation controls the dynamic balance of dislocation multiplication and annihilation, the dislocation interaction rule defines the physical process of dislocation intersection and entanglement, and the thermal activation energy threshold determines the energy required for dislocations to overcome obstacles under a certain stress level. The core of the micro-crack propagation module is the crack propagation rate formula and the fatigue life model under cyclic loading. The crack propagation rate formula establishes the relationship between the stress intensity factor range and the crack propagation speed, and the fatigue life model integrates the damage accumulation under the entire load history.

[0091] The calculation results of the three modules are output as intermediate data fields. The lattice distortion module outputs a lattice strain energy density cloud map, the dislocation motion module outputs a local energy dissipation rate distribution map, and the micro-crack propagation module outputs a crack driving force distribution map. The integration process performs pixel-level weighted superposition of the lattice strain energy density cloud map, the local energy dissipation rate distribution map and the crack driving force distribution map through a data fusion algorithm. The weighted coefficients are allocated according to the contribution of each module to the macro-mechanical response, generating a potential defect distribution map that comprehensively reflects the micro-damage evolution process of the material. In the potential defect distribution map, an automatic defect region recognition and labeling process needs to be performed. The recognition algorithm extracts the connected regions with stress concentration coefficients exceeding the critical value of the material based on image segmentation technology. The critical value is derived from the yield strength or fatigue limit data of the material, and these regions are labeled as high-risk defect regions. The coordinate information, geometric features and mechanical features of the high-risk defect regions are extracted and stored in a structured defect feature list. The operation of the thin-walled nickel alloy pipe material performance degradation evaluation model requires a high-performance computing environment with parallel computing units and large-capacity memory for processing large-scale finite element grids and transient dynamics calculations. The model parameter calibration stage uses known material typical damage experimental data to identify the key parameters in the lattice distortion module, dislocation motion module and micro-crack propagation module, ensuring the consistency of the model prediction results with the physical reality.

[0092] Example 3: see Figure 3, the water pressure loading strategy adjustment instruction is generated according to the potential defect distribution atlas, and the atlas needs to be analyzed for spatial characteristics, the spatial distribution characteristics and stress concentration degree of the high-risk defect area in the potential defect distribution atlas are analyzed, the spatial distribution characteristics include the aggregation degree, mutual distance and azimuth orientation of the defect area, and the stress concentration degree is quantified by calculating the ratio of the maximum stress concentration coefficient to the average stress concentration coefficient of each area. A stepped pressure increase instruction is generated for the identified intensive defect area, the stepped pressure increase instruction sets multiple intermediate pressure verification nodes, the pressure values of the intermediate pressure verification nodes increase in geometric progression, each intermediate pressure verification node maintains a fixed pressure holding time for observing the deformation response. A pulse loading instruction is generated for the isolated defect area in the graph, the pulse loading instruction adopts a short-time high-pressure impact mode, the peak pressure of the high-pressure impact is set to a certain multiple of the rated pressure, and the duration of the high-pressure impact is controlled in the order of milliseconds to stimulate the transient dynamics response of the defect. The pressure holding time parameter is adjusted according to the spatial density distribution of the defect area, the area with high spatial density distribution corresponds to a longer pressure holding time, and the adjustment of the pressure holding time follows a linear interpolation function based on the defect density. The stepped pressure increase instruction, the pulse loading instruction and the pressure holding time adjustment parameter are integrated into the water pressure loading strategy adjustment instruction set, the water pressure loading strategy adjustment instruction set adopts a structured data format to record the type, target area coordinates, pressure parameter sequence and time parameter of each instruction. The generation logic of the water pressure loading strategy adjustment instruction set is based on a defect response sensitivity model, which calculates the loading parameter of each grid area by the following formula:

[0093] ;

[0094] wherein: represents the adjusted pressure reference value of the i th grid area, represents the reference pressure value of the standard water pressure test, is a dimensionless stress sensitivity coefficient, represents the maximum equivalent stress value of the i th grid area, represents the yield strength of the thin-walled nickel alloy material, is a dimensionless defect aggregation coefficient, represents the total projection area of the defect cluster in the i th grid area, represents the total surface area of the measured area of the thin-walled nickel alloy pipe.

[0095] The micro-deformation sequence image captured by the high-frequency imaging unit is obtained by deploying a high-speed digital image correlation measurement system on the outer surface of the thin-walled nickel alloy pipe. The high-speed digital image correlation measurement system includes two high-resolution high-speed cameras and a speckle pattern preparation system, and the sampling resolution reaches the micron level. The system synchronously triggers three-dimensional digital image acquisition according to the pressure node set by the water pressure loading strategy adjustment instruction set. The synchronous trigger signal is obtained by hardware-level synchronization of the electrical signal sent by the pressure sensor and the camera exposure signal. Twenty frames of full-field strain distribution images are continuously captured at each pressure node, and the frame interval is not greater than five milliseconds to ensure that the strain propagation process can be analyzed. The image acquisition process is carried out in a stable lighting environment to reduce noise. The sub-pixel displacement tracking algorithm is used to process the sequence images. The sub-pixel displacement tracking algorithm reconstructs the time-varying process of the three-dimensional displacement field of the pipe wall surface by calculating the peak shift of the cross-correlation function of the speckle point group, and the displacement field calculation accuracy reaches the sub-pixel level. Three feature image layers, i.e., the maximum principal strain distribution, the minimum principal strain distribution and the shear strain distribution, are extracted from the three-dimensional displacement field based on the eigenvalue decomposition principle of the strain tensor. The feature image layers are arranged in order of pressure gradient to construct a micro-deformation sequence image database with spatio-temporal continuity. The micro-deformation sequence image database is stored in a four-dimensional tensor structure, and the four dimensions correspond to the spatial coordinates x, the spatial coordinates y, the pressure level p and the time frame t, respectively. The construction of the micro-deformation sequence image database completes the conversion from physical deformation to digital image, and provides high-fidelity experimental observation data for subsequent model verification.

[0096] In the water pressure test detection method for the inner wall of the thin-walled nickel alloy pipe, the difference comparison between the micro-deformation sequence image and the prediction result of the material performance degradation evaluation model is as follows: the maximum principal strain distribution image is extracted from the micro-deformation sequence image database as the actual observation data set. The maximum principal strain distribution image contains the actual strain value of each pixel point under water pressure load. The theoretical strain distribution graph is derived from the potential defect distribution graph output by the material performance degradation evaluation model after running. The theoretical strain distribution graph reflects the expected strain field calculated by the model based on the material micro-mechanical principle. The pixel-level spatial correspondence between the actual observation data set and the theoretical strain distribution graph is established, and the spatial correspondence is realized by the image registration algorithm, which unifies the two images to the same coordinate system and pixel scale. Based on the pixel-level correspondence, the strain value difference of each same position pixel is calculated to generate a local strain difference field covering the entire observation area of the thin-walled nickel alloy pipe. The local strain difference field stores the deviation amount of the prediction and the measurement in the form of a matrix.

[0097] The local strain difference field is subjected to regional cluster analysis. The regional cluster analysis adopts a density-based spatial clustering algorithm to identify abnormal response regions whose difference amplitudes exceed a preset standard deviation. The abnormal response regions are connected pixel point sets in space. Three key difference indicators of each identified abnormal response region are counted. The key difference indicators include an area ratio of the abnormal response region, a geometric shape factor of the abnormal response region, and a spatial distribution dispersion of the abnormal response region. The area ratio is a percentage of the area of the abnormal response region to the total observation area of the thin-walled nickel alloy pipe. The geometric shape factor describes the irregularity and compactness of the boundary of the abnormal response region. The spatial distribution dispersion measures the dispersion degree of multiple abnormal response regions in space. These key difference indicators are systematically recorded in a structured table for quantifying the deviation degree of the model prediction. See Table 1.

[0098] Table 1: Key difference indicators in the model prediction deviation quantification report

[0099] ;

[0100] A model prediction deviation quantification report is generated according to the calculated key difference indicators. The model prediction deviation quantification report records the location coordinates, size information, and specific values of the three key difference indicators of each abnormal response region in detail in the form of a document. The model prediction deviation quantification report clearly marks the defect identification parameters that need to be corrected, including the elastic constant tensor of the lattice distortion module, the critical shear stress threshold of the dislocation motion module, and the crack initiation criterion of the micro-crack propagation module. According to the deviation amount, the material performance degradation evaluation model needs to be iteratively updated by analyzing the defect identification parameter correction requirements marked in the model prediction deviation quantification report. The defect identification parameter correction requirements are directly related to the three core modules of the material performance degradation evaluation model. Adjust the elastic constant tensor parameter of the lattice distortion module. The adjustment of the elastic constant tensor parameter is based on the strain gradient distribution characteristics of the actual observation data set, and is realized by minimizing the root mean square error between the theoretical strain gradient and the measured strain gradient. Correct the critical shear stress threshold of the dislocation motion module. The correction of the critical shear stress threshold is based on the geometric shape factor of the abnormal response region. The abnormal region with a regular shape corresponds to dislocation sliding dominated deformation, and the abnormal region with an irregular shape corresponds to dislocation entanglement or grain boundary cracking. Optimize the crack initiation criterion of the micro-crack propagation module. The optimization of the crack initiation criterion is based on the spatial distribution dispersion of the abnormal response region. High dispersion indicates that crack initiation is random, and the statistical distribution parameter in the stress intensity factor calculation formula needs to be adjusted.

[0101] The three module parameters of the lattice distortion module, dislocation motion module and micro-crack propagation module are optimized in coordination. The gradient descent algorithm is used to iteratively adjust the module parameters, so that the correlation coefficient between the theoretical strain distribution map output by the model and the actual observation data set is improved to above the predetermined value. The correlation coefficient predetermined value is set to 0.95, indicating that the model prediction and experimental observation have a high degree of consistency. The optimized parameter configuration is stored in the configuration file of the model, including the updated elastic constant tensor, the revised critical shear stress threshold and the new crack initiation criterion. A new set of defect detection threshold range is generated, which is applied to the final hydrostatic test verification stage. The defect detection threshold range defines the qualified upper and lower limits of strain, stress and its rate of change. The completion of the iterative update of the material performance degradation evaluation model parameters marks the completion of the self-correction of the model based on the measured data, and improves the quantitative identification accuracy of the thin-walled nickel alloy pipe defect detection.

[0102] In the final verification stage of the thin-walled nickel alloy pipe inner wall hydrostatic test detection method, the optimized defect detection threshold range is used to perform the final hydrostatic test verification on the thin-walled nickel alloy pipe. The optimized defect detection threshold range defines the qualified upper and lower limits of axial strain, circumferential stress and radial displacement. Taking a thin-walled nickel alloy pipe numbered TN-8842 with a length of 8 meters as an example, a pressure maintaining test is performed at 1.5 times the rated working pressure of the thin-walled nickel alloy pipe, i.e. 45 MPa. The duration of the entire pressure maintaining process is set to 10 minutes to fully observe the creep behavior of the material. During the pressure maintaining test, the distributed sensing array monitors the deformation response data of the thin-walled nickel alloy pipe wall in real time at a sampling frequency of 1 kHz. The real-time monitoring data includes axial strain waveform, circumferential stress waveform and radial displacement waveform. The monitoring system compares the real-time waveform data with the optimized defect detection threshold range point by point. The comparison algorithm checks whether the data of the three channels is simultaneously within the threshold range at each sampling point.

[0103] When the monitoring system finds that the deformation response data of a certain area continuously exceeds the optimized defect detection threshold range for 100 sampling points (corresponding to a duration of 0.1 seconds), the system immediately automatically marks the center point coordinate position of the area and records the parameter type, the exceeding amplitude and the duration of the exceeding. For example, when the pressure test is carried out to the fourth minute, the sensor node located in the axial 3.2 meters and the circumferential 125-degree area of the thin-walled nickel alloy pipe detects that the circumferential stress value reaches 152 MPa, which exceeds the upper limit value 148 MPa of the optimized defect detection threshold range, and the system triggers an alarm and records the event. After the test is completed, the data processing software automatically summarizes all the recorded exceeding area information, generates a detailed log file containing the spatial position coordinates of the exceeding area, the exceeding parameter type, the maximum exceeding amplitude and the cumulative exceeding duration. According to the spatial distribution characteristics of the exceeding area and the parameter deviation degree, the system performs a pipe segment level automatic division algorithm, which divides the thin-walled nickel alloy pipe into three levels of qualified pipe segment, repairable pipe segment and scrap pipe segment. The judgment standard of the qualified pipe segment is that there is no any exceeding record in the whole test process, and all the deformation response curves are smooth and continuous. The judgment standard of the repairable pipe segment is that there is a local exceeding area, but the exceeding amplitude does not exceed 15% of the threshold range, and the exceeding area does not form a continuous strip distribution. The judgment standard of the scrap pipe segment is that the exceeding amplitude exceeds 15% of the threshold range, or the exceeding area forms a penetrating strip distribution, or multiple exceeding areas are densely distributed, resulting in an area ratio of more than 5%. Taking TN-8842 pipe as an example, the detection system finds that the circumferential stress exceeds 12% in the axial 0.5-0.8 meter area, and the radial displacement exceeds 8% in the axial 5.6-5.9 meter area, and these two areas are divided into repairable pipe segments; the circumferential stress exceeds 22% and the axial strain exceeds 18% in the axial 3.1-3.4 meter area, and this area is divided into a scrap pipe segment; the remaining non-exceeding area is divided into a qualified pipe segment.

[0104] The final detection report generation module outputs a final detection report containing the pipe section grade mark and defect coordinate list, and the final detection report adopts a structured data format, including pipe section basic information, detection condition summary, out-of-limit area detailed list and overall quality conclusion. The pipe section basic information records the number, size and material batch of the thin-walled nickel alloy pipe; the detection condition summary records the pressure curve, pressure holding time and environmental temperature of the final water pressure test; the out-of-limit area detailed list lists the starting coordinates, ending coordinates, out-of-limit parameter, maximum out-of-limit value and grade determination of each out-of-limit area in the form of a table; and the overall quality conclusion gives the overall qualified rate of the thin-walled nickel alloy pipe, repairable suggestion and scrap processing opinion. For the TN-8842 pipe, the final detection report clearly marks the length of the qualified pipe section as 6.2 meters, the length of the repairable pipe section as 0.6 meters, and the length of the scrap pipe section as 0.3 meters, and marks the area distribution with different colors in the attached figure. When the thin-walled nickel alloy pipe inner wall water pressure test detection system executes the computer program, the optimized parameters stored in the memory are called by the processor, the water pressure loading device is controlled to pressurize according to the set curve, and the sensor array and the data acquisition unit are coordinated to work synchronously. The processor compares and analyzes the real-time collected data with the optimized defect detection threshold range, and records the timestamp and coordinate information immediately when the out-of-limit condition is found. After the test is completed, the processor automatically executes the pipe section grade division algorithm, generates the final detection report and stores it in the specified path of the memory, and sends the report to the quality management department terminal.

[0105] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0106] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, replacements and changes can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for detecting thin-walled nickel alloy tube inner wall hydrostatic test, characterized in that, The method comprises the following steps: A closed water pressure environment is established inside the thin-walled nickel alloy pipe, and a preset pressure gradient is applied, and the pipe wall deformation response data is collected by a distributed sensing array; Multi-dimensional feature extraction is performed on the pipe wall deformation response data to generate a composite feature matrix containing axial strain features, circumferential stress features and radial displacement features; A thin-walled nickel alloy pipe material performance degradation evaluation model is constructed, the composite feature matrix is input into the model to simulate material microstructure changes, and a potential defect distribution map is output; A water pressure loading strategy adjustment instruction is generated according to the potential defect distribution map, and the pressure gradient and pressure holding time parameters in the water pressure environment are dynamically adjusted; A second water pressure test is performed based on the adjusted water pressure parameters, and a high-frequency imaging unit is used to capture pipe wall micro-deformation sequence images; The micro-deformation sequence images are compared with the prediction results of the material performance degradation evaluation model to identify the deviation between the actual defect area and the predicted defect area; The material performance degradation evaluation model is iteratively updated according to the deviation to generate an optimized defect detection threshold range; The thin-walled nickel alloy pipe is subjected to a final water pressure test verification using the optimized defect detection threshold range, and a qualified pipe segment marker and a defect pipe segment positioning coordinate are output; The multi-dimensional feature extraction on the pipe wall deformation response data comprises: Separating the axial strain waveform, the circumferential stress waveform and the radial displacement waveform from the pipe wall deformation response data set into three independent data channels; Extracting time domain statistical features, frequency domain energy features and nonlinear dynamic features from each data channel respectively; Using the time domain statistical features to construct a strain-stress correlation feature vector, including a peak factor, a waveform index and a pulse index; Using the frequency domain energy features to construct a resonance frequency band feature vector, including a dominant frequency component, a harmonic energy ratio and a frequency band power spectral density; Converting the nonlinear dynamic features into phase space reconstruction parameters to generate a dynamic feature vector representing the nonlinear response of the material; Fusing the strain-stress correlation feature vector, the resonance frequency band feature vector and the dynamic feature vector according to the time window to form a composite feature matrix; The construction of the thin-walled nickel alloy pipe material performance degradation evaluation model comprises: Establishing a multi-scale material response simulation framework including a lattice distortion module, a dislocation motion module and a micro-crack propagation module; Inputting the strain-stress correlation feature vector in the composite feature matrix into the lattice distortion module to calculate the theoretical lattice strain energy density distribution; Inputting the resonance frequency band feature vector into the dislocation motion module to simulate the local energy dissipation mode caused by the dislocation pinning effect; Driving the micro-crack propagation module to run by the dynamic feature vector to predict the crack tip stress field intensity factor change trajectory; Integrating the output results of the three modules to generate a potential defect distribution map reflecting the material micro-damage evolution process; Marking the area with a stress concentration coefficient exceeding a critical value as a high-risk defect area in the potential defect distribution map; The generation of the water pressure loading strategy adjustment instruction according to the potential defect distribution map comprises: Analyzing the spatial distribution characteristics and stress concentration degree of the high-risk defect area in the potential defect distribution map; Generating a stepped pressure increasing instruction for the dense defect area and setting multiple intermediate pressure verification nodes; The pulse loading instruction is generated for the isolated defect area, and a short-time high-pressure impact mode is used to stimulate the defect response; The pressure maintaining time parameter is adjusted according to the spatial density distribution of the defect area, and the pressure maintaining time is prolonged in the high-density area; The stepped pressure increasing instruction, the pulse loading instruction and the pressure maintaining time adjustment parameter are integrated into a water pressure loading strategy adjustment instruction set.

2. The method of claim 1, wherein the method further comprises, The pipe wall deformation response data is collected by the distributed sensing array, including: Three groups of optical fiber strain sensor groups are arranged axially on the thin-walled nickel alloy pipe, each group containing twelve equally spaced sensing nodes; All the sensing nodes are synchronously activated, and the initial water pressure loading stage pipe wall basic deformation data is recorded, with a sampling frequency not less than 1 kHz; The dynamic deformation response waveform is collected in the water pressure gradient rising stage, and the stable holding time of each pressure gradient step is recorded; The dynamic deformation response waveform is processed in time domain to eliminate the baseline drift caused by the difference in the installation position of the sensor; The processed dynamic deformation response waveform is spatio-temporally correlated with the corresponding pressure gradient value to generate a pipe wall deformation response data set with a pressure-deformation mapping relationship.

3. The method of claim 2, wherein the method further comprises, The pipe wall micro-deformation sequence image is captured by the high-frequency imaging unit, including: A high-speed digital image correlation measurement system is deployed on the outer surface of the thin-walled nickel alloy pipe, with a sampling resolution reaching the micron level; The three-dimensional digital image acquisition is synchronously triggered according to the pressure nodes set by the water pressure loading strategy adjustment instruction set; Twenty frames of full-field strain distribution images are continuously captured for each pressure node, with a frame interval not greater than five milliseconds; The sub-pixel displacement tracking algorithm is used to process the sequence images to reconstruct the time-varying process of the pipe wall surface three-dimensional displacement field; Three feature image layers, i.e., the maximum principal strain distribution, the minimum principal strain distribution and the shear strain distribution, are extracted from the three-dimensional displacement field; The feature image layers are arranged in the order of pressure gradient to construct a micro-deformation sequence image database with spatio-temporal continuity.

4. The method of claim 3, wherein the method further comprises, The micro-deformation sequence image is compared with the prediction result of the material performance degradation evaluation model, including: The maximum principal strain distribution image is extracted from the micro-deformation sequence image database as an actual observation data set; The theoretical strain distribution graph is derived from the potential defect distribution graph output by the material performance degradation evaluation model; The pixel-level correspondence between the actual observation data set and the theoretical strain distribution graph is established to calculate the local strain difference field; The local strain difference field is subjected to regional clustering analysis to identify abnormal response areas with a difference amplitude exceeding the standard deviation; Three key difference indicators, i.e., the area ratio, the geometric shape factor and the spatial distribution dispersion of the abnormal response area, are calculated; A model prediction deviation quantization report is generated according to the key difference indicators to mark the defect identification parameters that need to be corrected.

5. The method of claim 4, wherein the method further comprises, The parameter iterative update of the material performance degradation evaluation model according to the deviation amount includes: The defect identification parameter correction requirement marked in the model prediction deviation quantization report is analyzed; The elastic constant tensor parameter of the lattice distortion module is adjusted to match the strain gradient distribution of the actual observation data set; The critical shear stress threshold of the dislocation motion module is corrected based on the geometric shape factor of the abnormal response area. The crack initiation criterion of the micro crack propagation module is optimized, and the stress intensity factor calculation formula is updated according to the spatial distribution dispersion; Through the collaborative optimization of the three modules, the correlation coefficient between the theoretical strain distribution map output by the model and the actual observation data set is improved to above the predetermined value; The optimized parameters are configured and stored, and a new defect detection threshold range is generated and applied to the final hydrostatic test verification stage.

6. The method of claim 5, wherein the method further comprises, The final hydrostatic test verification of the thin-walled nickel alloy pipe using the optimized defect detection threshold range includes: Implementing a pressure test under the condition of one and a half times the rated working pressure of the thin-walled nickel alloy pipe, with a duration of not less than ten minutes; Real-time monitoring of the pipe wall deformation response data and point-by-point comparison with the optimized defect detection threshold range; When the deformation response of a certain area continuously exceeds the threshold range, immediately mark the coordinates of the area and record the overrun parameters; After the test, the spatial position information and parameter deviation of all overrun areas are summarized; According to the distribution characteristics of the overrun areas, the thin-walled nickel alloy pipe is divided into three grades: qualified pipe section, repairable pipe section and scrap pipe section; Output the final detection report containing the pipe section grade mark and defect coordinate list.

7. A thin walled nickel alloy tube inner wall hydrotest inspection system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein, The processor, when executing the computer program, realizes the steps of the thin-walled nickel alloy pipe inner wall hydrostatic test detection method in any one of claims 1 to 6.

Images