A method of final inspection of an aircraft turbine blade
Patent Information
- Application Number
- CN202611018484.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-09
- Publication Date
- 2026-09-08
AI Technical Summary
[0005]本发明的目的在于：提供了一种飞机涡轮叶片的成品检测方法，解决了传统单一物理场检测对叶片内部微观缺陷不灵敏、误判率高的问题
[0020]1. 检测灵敏度与可靠性更高： 本方法不单独分析温度场或振动场信号的绝对变化，而是计算从机械能耗散到温升的“传递熵”这一信息论指标，来量化双物理场之间能量耦合通路的完整性与效率。内部缺陷对能量传递路径的阻断或扭曲，会直接体现在局部能量耦合强度的显著降低上，这种异常比单一信号的变化更明显、更本质，从而能检出传统方法无法识别的微小缺陷，并显著降低误判率。
Smart Images
Figure CN122709591A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of non-destructive intelligent quality inspection technology for aircraft turbine blades, and relates to a finished product inspection method for aircraft turbine blades. Background Technology
[0002] Aircraft turbine blades, especially single-crystal superalloy blades, are the most valuable and harshest hot-end components in aero-engines. The integrity of their internal crystal structure, such as the presence of microscopic defects like impurities, spots, and porosity, directly determines the mechanical properties and service life of the blades.
[0003] Existing blade inspection methods largely rely on single-physical-field detection, such as using only X-ray diffraction or industrial computed tomography to examine internal structures, or using only ultrasound to detect internal cracks. These single-detection methods have inherent physical limitations. X-rays are insensitive to certain misaligned impurities, while ultrasound suffers severe scattering and attenuation in complex surfaces and microstructures, resulting in a low signal-to-noise ratio. More importantly, single-physical-field detection methods typically only analyze anomalies in the signal itself (such as amplitude attenuation and phase abrupt changes), failing to extract deeper, more fundamental perturbation information caused by internal blade defects to the energy transfer and conversion relationships between different physical fields. When the defect size is small or its physical properties are not significantly different from the matrix material, the anomaly of a single response is often drowned out by background noise, leading to low detection rates and high false positive rates.
[0004] Therefore, there is an urgent need for a method that can extract defect information from the multi-physics coupling mechanism, and achieve highly sensitive and reliable automated detection of micro-defects inside turbine blades by quantifying the energy transfer anomalies between different physical field responses. Summary of the Invention
[0005] The purpose of this invention is to provide a finished product inspection method for aircraft turbine blades, which solves the problems of traditional single physical field inspection being insensitive to microscopic defects inside the blades and having a high false judgment rate.
[0006] The technical solution adopted in this invention is as follows:
[0007] This application provides a method for inspecting finished aircraft turbine blades. The method includes: acquiring the spatiotemporal matrix of the temperature field and the spatiotemporal matrix of the vibration velocity field of the blade under multi-physics field coupling excitation, aligned to the same spatial grid; calculating the energy dissipation rate sequence and temperature rise rate sequence of each grid point in the time domain based on the vibration velocity field spatiotemporal matrix and the temperature field spatiotemporal matrix; for each grid point, calculating the transfer entropy from energy dissipation to temperature rise under a first time delay based on its energy dissipation rate sequence and temperature rise rate sequence, as the measured energy coupling strength of the grid point, wherein the first time delay is the time delay corresponding to the peak value of the transfer entropy from energy dissipation to temperature rise of a standard blade; performing point-by-point deviation calculation on the measured energy coupling matrix composed of the measured energy coupling strengths of all grid points and a pre-stored standard energy coupling matrix, and generating a detection reliability score to characterize the overall internal defect degree of the blade under test based on the result of the deviation calculation; and determining the blade under test as a defective product when the detection reliability score is lower than a preset pass threshold.
[0008] Optionally, the step of calculating the energy dissipation rate sequence and temperature rise rate sequence of each grid point in the time domain based on the spatiotemporal matrix of the vibration velocity field and the spatiotemporal matrix of the temperature field specifically includes:
[0009] Any element in the energy dissipation rate sequence is the unit area dissipation power calculated based on the velocity gradient of the grid point and its neighborhood at the corresponding time in the vibration velocity field spatiotemporal matrix.
[0010] Any element in the temperature rise rate sequence is the first derivative of the temperature value of that grid point with respect to time at the corresponding moment in the temperature field spatiotemporal matrix.
[0011] Optionally, the measured energy coupling matrix, formed by the measured energy coupling strength of all grid points, is compared with a pre-stored standard energy coupling matrix using a point-by-point deviation calculation. Based on the result of the deviation calculation, a detection reliability score is generated to characterize the overall internal defect level of the tested blade, specifically including:
[0012] ;
[0013] in, To test the credibility score, The total number of grid points. The measured energy coupling strength at the i-th grid point, The standard energy coupling strength is the i-th grid point in the standard energy coupling matrix.
[0014] Optionally, the standard energy coupling matrix can be generated in the following ways:
[0015] Multiple known defect-free turbine blade samples are acquired, and the operation of the detection system is performed on each sample to obtain the sample energy coupling matrix of each sample;
[0016] The energy coupling strength values corresponding to the same grid point in the sample energy coupling matrix of all samples are statistically averaged, and the calculated average value is used as the standard energy coupling strength of that grid point, thus forming the standard energy coupling matrix.
[0017] Optionally, the multi-physics coupling excitation is to apply a heat flow excitation with a preset power change curve to the blade under test, and simultaneously apply a vibration excitation with a frequency continuously sweeping within a preset frequency band.
[0018] Optionally, the vibration excitation is applied through a piezoelectric vibrator fixedly connected to the root of the blade under test, and the heat flow excitation is applied through an array of infrared radiation heaters whose power is controlled according to a preset program.
[0019] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:
[0020] 1. Higher Detection Sensitivity and Reliability: This method does not analyze the absolute changes in temperature or vibration field signals in isolation. Instead, it calculates the "transfer entropy," an information-theoretic index, from mechanical energy dissipation to temperature rise, to quantify the integrity and efficiency of the energy coupling path between the two physical fields. Internal defects that obstruct or distort the energy transfer path directly manifest as a significant reduction in local energy coupling strength. This anomaly is more obvious and fundamental than changes in a single signal, thus enabling the detection of minute defects that traditional methods cannot identify and significantly reducing the false positive rate.
[0021] 2. Achieve Quantitative and Standardized Judgment: By constructing a global "detection reliability score," the energy coupling deviations of various localities are integrated into a dimensionless score value. Comparison with a preset pass / fail threshold enables objective and unified automated pass / fail judgment, eliminating the subjectivity of manual interpretation.
[0022] 3. Optimal detection conditions were identified: By introducing the peak characteristics of the transfer entropy of a standard blade under the "first time delay", this method can identify and lock the coupling time delay with the richest information and the strongest defect characterization ability as the detection parameter, optimize the detection signal-to-noise ratio, and make the detection process no longer blind.
[0023] 4. Non-destructive, non-contact, and suitable for automated production lines: The thermal imaging and laser vibration measurement technologies used are non-contact full-field measurement technologies. The inspection process is non-destructive and fast, and it is easy to integrate into the automated inspection line of blade production to achieve full inspection and improve the overall quality control level. Attached Figure Description
[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort, wherein:
[0025] Figure 1 This is a schematic diagram of a finished product inspection method for aircraft turbine blades. Detailed Implementation
[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention; that is, the described embodiments are merely some embodiments of the invention, and not all embodiments. The components of the embodiments of the invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.
[0027] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.
[0028] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0029] The features and performance of the present invention will be further described in detail below with reference to embodiments.
[0030] Example 1:
[0031] This embodiment details the principle and implementation framework of a finished product inspection method for aircraft turbine blades. The method is applied to an automated inspection system that integrates a multi-physics excitation module, a sensor acquisition module, and a data processing and judgment module. The method includes:
[0032] Step S100: Obtain the spatiotemporal matrix of the temperature field and the spatiotemporal matrix of the vibration velocity field of the blade under test, aligned to the same spatial grid under multi-physics field coupling excitation. The multi-physics field coupling excitation is to apply a heat flow excitation with a preset power change curve to the blade under test, and simultaneously apply a vibration excitation with a frequency continuously sweeping within a preset frequency band. The vibration excitation is applied through a piezoelectric exciter fixedly connected to the root of the blade under test, and the heat flow excitation is applied through an infrared radiation heater array whose power is controlled according to a preset program.
[0033] Step S200: Based on the spatiotemporal matrix of the vibration velocity field and the spatiotemporal matrix of the temperature field, calculate the energy dissipation rate sequence and temperature rise rate sequence of each grid point in the time domain, specifically including:
[0034] Any element in the energy dissipation rate sequence is the unit area dissipation power calculated based on the velocity gradient of the grid point and its neighborhood at the corresponding time in the vibration velocity field spatiotemporal matrix.
[0035] Any element in the temperature rise rate sequence is the first derivative of the temperature value of that grid point with respect to time at the corresponding moment in the temperature field spatiotemporal matrix.
[0036] Step S300: For each grid point, based on its energy dissipation rate sequence and temperature rise rate sequence, calculate the transfer entropy from energy dissipation to temperature rise under the first time delay, as the measured energy coupling strength of the grid point, wherein the first time delay is the time delay corresponding to the peak value of the transfer entropy from energy dissipation to temperature rise of the standard blade.
[0037] Step S400: The measured energy coupling matrix, composed of the measured energy coupling strengths of all grid points, is compared with the pre-stored standard energy coupling matrix by point-by-point deviation calculation. Based on the result of the deviation calculation, a detection reliability score is generated to characterize the overall internal defect level of the tested blade. The specific calculation method is as follows:
[0038] ;
[0039] in, To test the credibility score, The total number of grid points. The measured energy coupling strength at the i-th grid point, The standard energy coupling strength is the i-th grid point in the standard energy coupling matrix.
[0040] The standard energy coupling matrix is generated in the following ways:
[0041] Multiple known defect-free turbine blade samples are acquired, and the operation of the detection system is performed on each sample to obtain the sample energy coupling matrix of each sample;
[0042] The energy coupling strength values corresponding to the same grid point in the sample energy coupling matrix of all samples are statistically averaged, and the calculated average value is used as the standard energy coupling strength of that grid point, thus forming the standard energy coupling matrix.
[0043] Step S500: When the detection reliability score is lower than the preset pass threshold, the tested blade is determined to be a defective product.
[0044] The specific implementation method of the above steps can be as follows:
[0045] I. Obtaining Multiphysics Excitation and Spatiotemporal Matrix:
[0046] The detection system first applies a multi-physics coupled excitation program to the blade under test. Specifically, an infrared radiation heater array with power controlled according to a preset program applies a heat flux excitation with a specific spatiotemporal distribution to the blade surface; at the same time, a broadband vibration excitation with a frequency continuously sweeping within a preset frequency band is applied through a piezoelectric vibrator fixedly connected to the blade root.
[0047] Under this coupled excitation, the detection system operates synchronously:
[0048] 1. A high-resolution infrared thermal imager was used to continuously acquire a sequence of thermal images of the blade surface at a fixed frame rate.
[0049] 2. Using a scanning laser Doppler vibration meter, the sequence of out-of-surface vibration velocity maps of each preset grid point on the blade surface is collected simultaneously.
[0050] After receiving the raw signals, the data processing and judgment module performs a spatiotemporal alignment operation: mapping and interpolating each frame of thermal image and vibration velocity image onto a predefined spatial grid that fits the three-dimensional surface of the blade, thereby constructing the spatiotemporal matrix of the temperature field and the spatiotemporal matrix of the vibration velocity field. Each element of the spatiotemporal matrix of the temperature field and the spatiotemporal matrix of the vibration velocity field represents the temperature value or vibration velocity value of a specific grid point at a specific time.
[0051] II. Calculation of Derived Intermediate Physical Quantity Fields:
[0052] After obtaining the spatiotemporal matrices of the temperature field and the vibration velocity field, the detection system begins to calculate two key intermediate physical quantity sequences:
[0053] Energy dissipation rate sequence: For any grid point, each element in the energy dissipation rate sequence is the mechanical energy dissipation power per unit area calculated using a viscoelastic dissipation model based on the velocity gradient of that grid point and its neighborhood at the corresponding time in the spatiotemporal matrix of the vibration velocity field. Arranging the energy dissipation rate values over the entire time history constitutes the energy dissipation rate sequence for that grid point. The energy dissipation rate sequence reflects the rate of change in the conversion of mechanical energy into heat energy.
[0054] Temperature rise rate sequence: For any grid point, each element in the temperature rise rate sequence is the first derivative of the temperature value at that grid point with respect to time in the temperature field spatiotemporal matrix at the corresponding moment. Arranging the temperature rise rate values over the entire time history constitutes the temperature rise rate sequence for that grid point. The temperature rise rate sequence reflects the rate of change in heat accumulation.
[0055] III. Extraction of core features: Calculation of measured energy coupling strength.
[0056] The core innovation of this method lies in calculating the transfer entropy from the energy dissipation rate sequence to the temperature rise rate sequence, rather than examining the energy dissipation rate sequence or the temperature rise rate sequence in isolation. Transfer entropy is an asymmetric statistic based on information theory. It quantifies the reduction in the uncertainty of predicting the future state of one time series (the source sequence) by the historical state of another time series (the target sequence), excluding the influence of the target sequence's own history. In this scenario, transfer entropy precisely measures the contribution of the activity history of the "cause"—"how much energy was dissipated mechanically"—to the prediction of the future behavior of the "effect"—"how much the temperature rose," i.e., the strength and smoothness of the energy coupling path between the two physics fields.
[0057] The detection system needs to calculate the transfer entropy under a specific time delay, which is defined as the first time delay. The first time delay is determined by pre-applying the same multiphysics coupling excitation program to multiple standard defect-free blades, calculating the transfer entropy from energy dissipation to temperature rise at each grid point under various time delays, and finding the time delay that causes the transfer entropy curve to reach its peak. This peak time delay represents the optimal delay time for the stress-thermal coupling response in the defect-free material, resulting in the highest energy transfer efficiency and the largest amount of information. The detection system uses this peak time delay as a standard parameter and pre-stores it as the first time delay.
[0058] During detection, for each grid point on the spatial grid, the detection system calculates the transfer entropy value from energy dissipation to temperature rise under the first time delay, based on the energy dissipation rate sequence and temperature rise rate sequence calculated for that grid point. This value is recorded as the measured energy coupling strength of that grid point. By traversing all grid points, all the calculated measured energy coupling strength values constitute a measured energy coupling matrix that corresponds one-to-one with the spatial grid.
[0059] IV. Deviation Quantification and Final Decision-Making.
[0060] Obtaining only the measured energy coupling matrix is insufficient to directly determine whether a blade is qualified, because even for defect-free blades, the absolute coupling strength at different locations (such as the fin, blade body, and tenon) may vary due to structural differences. Therefore, it is necessary to compare it with a standard template representing a "perfect" state.
[0061] The detection system has a pre-stored standard energy coupling matrix. The standard energy coupling matrix is generated as follows: multiple known defect-free turbine blade samples are obtained, and all the operations described above in this embodiment are performed on each turbine blade sample to obtain a sample energy coupling matrix for each turbine blade sample; then, the energy coupling intensity values corresponding to the same grid point in the sample energy coupling matrix of all turbine blade samples are statistically averaged, and the average value calculated for each grid point is used as the standard energy coupling intensity of that grid point, thus forming the standard energy coupling matrix.
[0062] Next, the detection system calculates the point-by-point deviation between the measured energy coupling matrix and the standard energy coupling matrix, and generates a final detection reliability score.
[0063] Specifically, the calculation uses a non-linear weighted scoring function:
[0064] ;
[0065] in, To test the credibility score, The total number of grid points. The measured energy coupling strength at the i-th grid point, The standard energy coupling strength is the i-th grid point in the standard energy coupling matrix.
[0066] The design principle of this function is:
[0067] It is a local bias penalty factor. When the measured energy coupling strength at a certain grid point i... Coupling strength with standard energy When all points are identical, the local deviation penalty factor is 1 (full marks). As the absolute value of the deviation increases relative to the standard energy coupling strength M_ref(i), the local deviation penalty factor decreases exponentially, tending towards 0. This achieves severe penalty for points that significantly deviate from the standard pattern.
[0068] This is the confidence weight of the local coupling strength. It ensures that the contribution of the grid point score to the overall score is related to its own energy coupling strength. A grid point with low coupling strength (which may be computationally unstable even without defects due to weak signals) naturally has a weakened impact on the final decision; while a deviation from a high coupling strength point will be amplified because it should be a "sure thing" anomaly.
[0069] Finally, by summing and dividing by the total number of grid points N, a normalization is achieved, resulting in a value between 0 and max( The reliability score between the detection and the test results.
[0070] Finally, the detection system compares the detection reliability score with a pre-determined pass / fail threshold determined through extensive experiments and statistical analysis. When the detection reliability score is below the pass / fail threshold, it indicates that there is a region inside the blade that significantly deviates from the standard energy coupling mode. The detection system then determines that the tested blade is defective and drives the sorting mechanism to separate the tested blade.
[0071] Example 2:
[0072] This embodiment, in conjunction with specific numerical values and parameters, details the complete process of performing finished product testing on a specific batch of high-pressure turbine blades using the above method.
[0073] 1. Establishment of the test object and standard template:
[0074] The tested object is a newly cast nickel-based single-crystal high-temperature alloy high-pressure turbine blade, with a batch number of VAN-2310 and a typical blade wall thickness of 1.5 to 3 mm.
[0075] Before implementing automated batch inspection, a standard model is established. Thirty blades of the same model that have been confirmed to be internally defect-free and have qualified crystal orientation through multiple methods (including high-sensitivity X-ray diffraction, industrial CT, and metallographic sampling) are selected as the standard sample group.
[0076] Meanwhile, five blades of the same model containing known artificial defects were prepared as a calibration sample group. The artificial defects were created by electrical discharge machining of microholes or grooves with a size range of 0.2 to 0.5 mm in different parts of the blades to simulate defects such as impurities and porosity that may occur in actual service.
[0077] 2. System parameter settings and multiphysics coupling excitation program:
[0078] The detection system adaptively divides the blade surface into approximately 10,000 spatial grid points based on the blade surface area, corresponding to a spatial resolution of approximately 50 points per square centimeter.
[0079] The multiphysics coupling excitation program is set as follows:
[0080] Thermal excitation: The infrared radiation heater array linearly increases from 0 to a peak power of 50 kW per square meter within 1 second, then maintains a constant power for 3 seconds, and then linearly decreases to 0 within 1 second.
[0081] Vibration excitation: A piezoelectric exciter applies a continuous sinusoidal sweep signal from 100 Hz to 5000 Hz with a sweep period of 2 seconds, cyclically applied throughout the heating process. The low-frequency band (100 to 2000 Hz) is primarily used to excite the overall blade modes to assess its macroscopic structural integrity; the high-frequency band (2000 to 5000 Hz), where the elastic wave wavelength has shortened to the centimeter level, begins to produce discernible scattering and energy dissipation effects on local stiffness changes and minute defects. The two frequency bands work synergistically, complementing the thermal excitation channel.
[0082] Acquisition settings and timing alignment:
[0083] The infrared thermal imager records continuously at a fixed frame rate of 100 Hz, that is, it obtains one frame of temperature distribution data of the whole field every 10 milliseconds.
[0084] Vibration information was acquired using a scanning laser Doppler vibrometer, scanning each preset grid point on the blade surface at a sampling rate of 50,000 Hz. After receiving the complete time-domain velocity signal from each measurement point, the data processing and judgment module reconstructed the signal: utilizing the high sampling rate to accurately capture the velocity waveform of each vibration cycle, it calculated the peak energy dissipation power caused by material damping and defects within each 10-millisecond time window. This peak power was used as the energy dissipation rate at that moment and for that grid point, thus constructing a spatiotemporal matrix of the energy dissipation rate field that is strictly aligned frame-by-frame with the infrared thermal imager. Therefore, the spatiotemporal matrix of the energy dissipation rate field has the exact same time dimension and time step as the spatiotemporal matrix of the temperature field, providing a reliable data foundation for subsequent transfer entropy calculations.
[0085] 3. Calibration of the first time delay and the standard energy coupling matrix:
[0086] The above excitation, acquisition, and timing alignment procedures were executed on each of the 30 standard sample blades. The data processing and judgment module calculated the energy dissipation rate sequence and temperature rise rate sequence for each blade and each grid point, and calculated the transfer entropy by scanning in 1-millisecond steps within the range of 0 to 100 milliseconds. For each grid point, the transfer entropy curves of all 30 blades were statistically averaged to find the time delay corresponding to the peak value of the average curve. The peak time delays of all grid points were counted, and the mode was taken as the first time delay. In this example, the first time delay was calculated to be 12.5 milliseconds.
[0087] The physical meaning of this 12.5 milliseconds can be interpreted as follows: it is approximately equal to the damping relaxation time excited by high-frequency vibration in the thin-walled structure of the blade, or the characteristic time of the local thermoelastic coupling response. Under the specific excitation and material combination of this detection system, the transfer entropy is the largest at this time delay, indicating that the information flow of the energy coupling path is clearest at this point, and the disturbance caused by the defect is most easily captured.
[0088] After determining the first time delay, the sample energy coupling matrix of each standard sample blade under that time delay is extracted. The energy coupling strength values of the same grid point in all sample energy coupling matrices are arithmetically averaged to obtain the final standard energy coupling matrix containing 10,000 value units.
[0089] 4. Calibration and determination of the acceptable threshold:
[0090] To ensure the reliability of the judgment criteria, the same testing procedure as described above was performed on the 5 calibration sample blades, and their respective test reliability scores were obtained. Simultaneously, scores were also calculated for the 30 standard sample blades. Statistical analysis was performed on the two sets of scores. A threshold was determined to achieve an optimal balance between the false rejection rate (false positive) for standard samples and the false negative rate (false negative) for calibration samples. Through receiver operating characteristic (ROC) curve analysis, the pass / fail threshold was determined to be 0.0400 in this example. At this threshold, the scores of all known defective samples were below 0.0400, while the scores of all defect-free samples were above 0.0400.
[0091] 5. Automated detection and judgment of the blade under test:
[0092] The finished blade with batch number VAN-2310-0457 is now being tested. The testing system automatically performs identical excitation, acquisition, and timing alignment processes to generate the temperature field spatiotemporal matrix and energy dissipation rate field spatiotemporal matrix for the blade. Then, the energy dissipation rate sequence and temperature rise rate sequence for each of the 10,000 grid points are calculated. Subsequently, the testing system calculates the transfer entropy of all grid points at the first time delay of 12.5 milliseconds, forming the measured energy coupling matrix of the blade. The testing system retrieves the pre-stored standard energy coupling matrix and substitutes it into the scoring function:
[0093] ;
[0094] After point-by-point calculation, summation, and normalization, the detection reliability score of the leaf was found to be 0.0312.
[0095] Since the score of 0.0312 < 0.0400, the detection system determines that the blade fails the inspection. The interface displays "Defective Product," a red warning light illuminates, and the automatic sorting robotic arm picks up the blade from the conveyor belt and places it into the waste recycling channel. Subsequently, the detection system automatically resets, ready to receive and inspect the next blade.
[0096] This paper proposes a multiphysics coupling analysis method for the inspection of finished aircraft turbine blades. Its core principle is not to directly explore the defects themselves, but to monitor and quantify the disturbances caused by the defects to the "energy transfer pathways" inside the blade.
[0097] Specifically, this technical solution follows the following technical logic chain:
[0098] 1. Apply coupling excitation to construct an energy transfer path:
[0099] Controllable thermal excitation and broadband vibration excitation are simultaneously applied to the blade under test. The combined effect of these two physical fields establishes a deterministic "mechanical energy-thermal energy" transfer and conversion channel inside the blade. In a defect-free ideal material, there is a stable and predictable spatiotemporal coupling relationship between local mechanical energy dissipation (internal friction caused by vibration) and the instantaneous temperature rise it causes.
[0100] 2. Extract information transfer indicators across physical fields:
[0101] By simultaneously acquiring high spatiotemporal resolution data of temperature and vibration velocity fields, the system calculates two key derived time series: the energy dissipation rate series and the temperature rise rate series. Then, the concept of transfer entropy from information theory is introduced to quantify the intensity of information transfer from "energy dissipation" to "temperature rise." The level of transfer entropy directly reflects the smoothness of the energy coupling path at that location.
[0102] 3. The blocking effect of mapping defects on energy pathways:
[0103] Microscopic defects inside the blade (such as impurities, porosity, and cracks) can severely disrupt this ideal energy transfer path. This manifests as the heat generated by localized energy dissipation failing to contribute effectively and promptly to the temperature rise, leading to an interruption or weakening of the causal chain of information transmission. This anomaly will be directly reflected in a significant reduction in the measured energy coupling strength at the measuring point.
[0104] 4. Compare with the "gold standard" to quantify the overall deviation:
[0105] A large amount of defect-free blade data was collected to establish a "standard energy coupling matrix" representing the ideal state. The "measured energy coupling matrix" of the blade under test was compared point by point with the standard matrix. A nonlinear scoring function was used to aggregate and amplify the small deviations at each point into a single, quantitative test reliability score.
[0106] 5. Achieving objective and highly sensitive automated judgment: This detection reliability score comprehensively reflects the degree to which the blade deviates from the ideal energy coupling mode. Automatic pass / fail determination of the blade can be achieved by using a preset pass / fail threshold. Because this method detects the deep disturbance of the energy transfer mechanism caused by defects, rather than the amplitude change of a single signal, its detection sensitivity and reliability for minute defects are superior to traditional single-physics field methods.
[0107] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for inspecting finished aircraft turbine blades, characterized in that, The method includes: Obtain the spatiotemporal matrices of the temperature field and vibration velocity field of the blade under test, aligned to the same spatial grid, under multi-physics coupled excitation. Based on the spatiotemporal matrix of the vibration velocity field and the spatiotemporal matrix of the temperature field, the energy dissipation rate sequence and the temperature rise rate sequence of each grid point in the time domain are calculated respectively. For each grid point, based on its energy dissipation rate sequence and temperature rise rate sequence, the transfer entropy from energy dissipation to temperature rise under the first time delay is calculated as the measured energy coupling strength of the grid point, wherein the first time delay is the time delay corresponding to when the transfer entropy from energy dissipation to temperature rise of the standard blade reaches its peak. The measured energy coupling matrix, which is composed of the measured energy coupling strength of all grid points, is compared with the pre-stored standard energy coupling matrix by point-by-point deviation calculation. Based on the result of the deviation calculation, a detection reliability score is generated to characterize the degree of internal defects of the tested blade as a whole. When the detection reliability score is lower than the preset pass threshold, the tested blade is determined to be a defective product.
2. The method for inspecting finished aircraft turbine blades according to claim 1, characterized in that, The calculation of the energy dissipation rate sequence and temperature rise rate sequence for each grid point in the time domain, based on the spatiotemporal matrix of the vibration velocity field and the spatiotemporal matrix of the temperature field, specifically includes: Any element in the energy dissipation rate sequence is the unit area dissipation power calculated based on the velocity gradient of the grid point and its neighborhood at the corresponding time in the vibration velocity field spatiotemporal matrix. Any element in the temperature rise rate sequence is the first derivative of the temperature value of that grid point with respect to time at the corresponding moment in the temperature field spatiotemporal matrix.
3. The finished product inspection method for aircraft turbine blades according to claim 1, characterized in that, The measured energy coupling matrix, formed by the measured energy coupling strength of all grid points, is compared with a pre-stored standard energy coupling matrix using a point-by-point deviation calculation. Based on the result of the deviation calculation, a detection reliability score is generated to characterize the overall internal defect level of the tested blade. Specifically, this includes: ； in, To test the credibility score, The total number of grid points. The measured energy coupling strength at the i-th grid point, The standard energy coupling strength is the i-th grid point in the standard energy coupling matrix.
4. A method for inspecting finished aircraft turbine blades according to any one of claims 1 and 3, characterized in that, The standard energy coupling matrix is generated in the following ways: Multiple known defect-free turbine blade samples are acquired, and the operation of the detection system is performed on each sample to obtain the sample energy coupling matrix of each sample; The energy coupling strength values corresponding to the same grid point in the sample energy coupling matrix of all samples are statistically averaged, and the calculated average value is used as the standard energy coupling strength of that grid point, thus forming the standard energy coupling matrix.
5. The finished product inspection method for aircraft turbine blades according to claim 1, characterized in that, The multi-physics coupling excitation is to apply a heat flow excitation with a preset power change curve to the blade under test, and at the same time apply a vibration excitation with a frequency continuously sweeping within a preset frequency band.
6. The finished product inspection method for aircraft turbine blades according to claim 5, characterized in that, The vibration excitation is applied through a piezoelectric vibrator fixedly connected to the root of the blade under test, and the heat flow excitation is applied through an array of infrared radiation heaters whose power is controlled according to a preset program.