Rotor-stator tip radius measurement method and system for an aeroengine

Abstract

The application discloses an aero-engine rotor-stator blade tip polar radius measuring method and system. The method comprises the following steps: using a measuring needle to perform layered scanning on the blade top at different heights perpendicular to the engine axis to obtain three-dimensional coordinate data (X, Y, Z) and normal vector data (I, J, K); establishing a polar coordinate system with the engine axis as the origin on the scanning height plane to convert the plane coordinates into polar coordinates (R, theta); determining the segmented polar angle based on the adaptive segmentation algorithm of the theoretical angle interval guidance and sector extreme value search, and segmenting the data into blade point cloud subsets; determining the blade tip vertex and calculating the distance from the vertex to the axis as the blade tip polar radius. The application has high segmentation accuracy and realizes high-precision and high-efficiency automatic measurement of the blade tip polar radius.

Description

Technical Field

[0001] This invention belongs to the field of aero-engine blade tip data measurement technology, specifically, it relates to a method and system for measuring the extreme radius of aero-engine rotor-stator blade tip. Background Technology

[0002] In aero-engine maintenance, the blade tip polar radius is a core geometric parameter characterizing the assembly accuracy and operational reliability of rotor and stator blades. Its measurement accuracy directly determines the engine's aerodynamic efficiency recovery and service life stability. Conducting precise measurement of the blade tip polar radius aims to verify whether the post-repair blade assembly position meets design tolerance requirements by quantifying the polar distance from the blade tip to the engine's rotation axis. It also helps identify polar radius deviations caused by wear, deformation, or assembly errors, providing crucial data support for blade rework, adjustment, and overall engine assembly acceptance.

[0003] Existing blade tip radius measurement technology has the following key drawbacks:

[0004] 1. Traditional measurement methods mostly use spherical probes for single-point contact measurement, which is cumbersome and inefficient. The sampling of the blade tip curve is discontinuous, and it is easy to miss key extreme points. In addition, the contact area between the spherical probe and the blade tip surface is small, and the measurement is sensitive to noise, which cannot meet the high-precision scanning requirements of the complex blade tip surface.

[0005] 2. Some technologies use non-contact measurement (such as laser scanning), but they are easily affected by reflections and oil stains on the blade surface, resulting in unstable measurement accuracy. In addition, the equipment cost is high, making it difficult to adapt to industrial batch testing scenarios.

[0006] 3. Even when using coordinate measuring machines (CMMs), data is mostly collected manually at single points, resulting in low accuracy and efficiency, which cannot meet the needs of mass production.

[0007] The information disclosed in this background section is only intended to enhance the understanding of the background technology of this application, and therefore may include prior art that is not known to those skilled in the art. Summary of the Invention

[0008] This invention proposes a method and system for measuring the tip radius of the rotor stator blades of aero-engines, in order to solve the technical problems of low accuracy and low efficiency of existing contact measurement methods for single-point data acquisition.

[0009] To achieve the above-mentioned invention / design objectives, the present invention adopts the following technical solution:

[0010] A method for measuring the tip radius of a rotor stator blade in an aero-engine, the method comprising:

[0011] The probe was used to perform layered scanning of the blade tips at different heights perpendicular to the engine axis to obtain the three-dimensional coordinate data (X, Y, Z) of the blade tip at each height and the normal vector data (I, J, K) of the contact point.

[0012] In each scanning height plane, a polar coordinate system is established with the engine axis as the origin, and the planar coordinates in the three-dimensional coordinate system are taken. Convert to polar coordinate data (polar radius R, polar angle θ);

[0013] The segmented polar angle is determined based on the polar radius R and the polar angle θ. The polar coordinate data is then segmented based on the segmented polar angle, and each segmentation results form a subset of point cloud.

[0014] For each subset of point clouds, determine the blade tip vertex and calculate the distance from each blade tip vertex to the engine axis as the blade tip polar radius.

[0015] The method for measuring the tip polar radius of aero-engine stator blades as described above, which determines the segmented polar angle based on the polar radius R and polar angle θ, includes:

[0016] Determine the sampling point with the largest polar radius The corresponding polar angle is ;

[0017] extract Preorder neighbor Adjacent points to the next order The corresponding polar angles are respectively and ;

[0018] calculate and , The absolute difference is used to select the neighboring points corresponding to the larger difference as... The corresponding polar angle is ;

[0019] calculate and The average value of the polar angles yields the first segmented polar angle. ;

[0020] The theoretical angular interval corresponding to a single blade is , Total number of blades;

[0021] Segmentation and splitting angle update, , obtained the Theoretical search perspectives for each segment;

[0022] Determine the angle sector, in order to Centered on, delineate the first Each segment of angle sector ;

[0023] In sector The point corresponding to the maximum value of the inner search polar radius ;

[0024] extract Preorder neighbor Adjacent points to the next order The corresponding polar angles are respectively and ;

[0025] calculate and The absolute difference is used to select the neighboring points corresponding to the larger difference as... The corresponding polar angle is ;

[0026] calculate and The average value of the polar angles is used to obtain the i-th segment polar angle. ;

[0027] The coordinate data is segmented based on the segmented polar angle.

[0028] The above-described method for measuring the tip radius of aero-engine stator blades involves determining the angular sector and then normalizing the sector. Then the sector is split into ;like Then the sector is split into ;

[0029] For in sectors or The point corresponding to the maximum value of the inner search polar radius .

[0030] The above-described method for measuring the tip polar radius of aero-engine stator blades applies to all segmented polar angles. Normalization calibration is performed to form a normalized and calibrated segmented point cloud subset. For each segmented point cloud subset, outliers are detected and removed based on iterative residual statistical analysis and adaptive spline fitting, using a dual robust statistical discrimination criterion, which is a dual discrimination fusion of median absolute deviation and standard deviation.

[0031] The method for measuring the tip radius of the rotor stator blades of an aero-engine, as described above, has the blade installation angle as follows: ;

[0032] Compensation amount R is the radius of the probe's arc surface;

[0033] Correcting the leaf tip vertex coordinates:

[0034] ;

[0035] ;

[0036] By calculating the distance from the blade tip to the engine axis .

[0037] The method for measuring the tip radius of the rotor stator blades of an aero-engine, as described above, involves the diameter of the probe. , This is the maximum gap between the tips of adjacent leaf blades.

[0038] As described above, the method for measuring the tip radius of aero-engine stator blades involves touching the calibration area at the tip of blade number one and recording the three-dimensional coordinates of that contact point. The corresponding polar angle is θ l0 .

[0039] The above-described method for measuring the polar radius of aero-engine rotor-stator blade tips calculates the polar angle at each blade tip apex. With θ l0 The difference is calculated, and the segment corresponding to the tip of the leaf with the smallest difference is leaf number 1. The remaining segments are numbered sequentially from 2 to N in a clockwise or counterclockwise direction.

[0040] As described above, in the method for measuring the tip polar radius of aero-engine stator blades, the tip apex number calculated at other altitudes within the same level is based on the polar angle θ calibrated at the first blade contact point of that level. l0 Matching is performed based on the baseline.

[0041] A system for measuring the tip radius of a rotor stator blade in an aero-engine, the system comprising a probe and a data processing system, wherein the data processing system performs the measurement according to the aforementioned measurement method.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are as follows: The method for measuring the tip polar radius of the rotor stator blade of the present invention includes: using a probe to perform layered scanning of the blade tip at different heights perpendicular to the engine axis, obtaining the three-dimensional coordinate data (X, Y, Z) of the blade tip at each height and the normal vector data (I, J, K) of the contact point; establishing a polar coordinate system with the engine axis as the origin in each scanning height plane, and taking the planar coordinates in the three-dimensional coordinate system. The data is converted into polar coordinates (polar radius R, polar angle θ). Based on the polar radius R and polar angle θ, segmented polar angles are determined. The polar coordinate data is then segmented according to these segmented polar angles, with each segment forming a subset of point clouds. For each subset of point clouds, the blade tip vertex is determined, and the distance from each blade tip vertex to the engine axis is calculated as the blade tip polar radius. This invention achieves high segmentation accuracy and enables high-precision, high-efficiency measurement of the blade tip polar radius.

[0043] Other features and advantages of the present invention will become clearer after reading the detailed embodiments of the invention in conjunction with the accompanying drawings. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a schematic diagram of a disc-shaped probe according to a specific embodiment of the present invention.

[0046] Figure 2 This is a schematic diagram of a single-layer stator blade according to a specific embodiment of the present invention.

[0047] Figure 3 This is a schematic diagram of a single-layer rotor blade according to a specific embodiment of the present invention.

[0048] Figure 4 This is a scanning data of a certain height of the stator blade of a single-layer casing in a specific embodiment of the present invention.

[0049] Figure 5 yes Figure 4 Enlarged view of the area within the box.

[0050] Figure 6 This is a schematic diagram of segmented polar angles in a specific embodiment of the present invention.

[0051] Figure 7 This is a schematic diagram of the cosine error compensation for the probe in a specific embodiment of the present invention.

[0052] Figure 8 This is a flowchart of a specific embodiment of the present invention. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] In the description of this invention, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0055] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. In the description of embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0056] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature.

[0057] In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0058] This embodiment solves the problems in existing methods for measuring the tip radius of rotor blades in aero-engines, such as discontinuous sampling, inaccurate segmentation of the tip curve, low degree of automation, cumbersome and inefficient operation, limited accuracy, insufficient error compensation, easy for the probe to fall into the blade gap leading to measurement failure, and inability to accurately associate blade numbers.

[0059] This embodiment combines CMM with a disc probe (such as...) Figure 1 Based on the above, the following improvement scheme is proposed:

[0060] 1. The selection principle is that the probe diameter should be greater than the maximum gap between the tips of adjacent blades.

[0061] In some embodiments, the probe is a disc-shaped probe or a spherical probe.

[0062] 2. Add a step to calibrate the first blade contact point to achieve accurate matching of blade numbers.

[0063] 3. Continuous blade tip curve data are acquired through scanning. An adaptive search segmentation algorithm based on angle intervals is used to solve the curve segmentation problem. Error compensation and data processing algorithms adapted to the characteristics of the probe are designed to form an integrated, high-precision, high-reliability, and traceable measurement solution.

[0064] The core innovation of the aero-engine stator blade tip radius measurement method in this embodiment lies in the use of CMM with a large-diameter probe for continuous scanning sampling of the blade tip curve, the clear selection principle of probe diameter being greater than the maximum blade clearance, the addition of a first blade contact point for calibration to achieve numbering traceability, the introduction of an adaptive search segmentation algorithm based on angle intervals to optimize the segmentation of scanning data, and the design of a full-process processing scheme based on this sampling method, adaptation principle, positioning mechanism and segmentation algorithm.

[0065] like Figure 2 The diagram shown is a schematic of a single-layer stator blade. In some embodiments, the engine may also include multiple layers of stator blades.

[0066] like Figure 3 The diagram shown is of a single-layer rotor blade. In some embodiments, the engine may also include multiple layers of rotor blades.

[0067] like Figure 8 As shown, the specific steps include:

[0068] 1. Measurement system setup and parameter configuration:

[0069] 1) Core measuring equipment: CMM + large diameter probe (disc probe, spherical probe, etc.). The disc probe's disc end face contacts the blade tip curved surface to achieve continuous scanning sampling and establish a coordinate system based on the drawings.

[0070] 2) Probe selection: Based on the design parameters or measured data of the engine blade to be measured, determine the maximum clearance between the tips of adjacent blades. (Unit: mm), Select the diameter of the disc probe. (Unit: mm) Satisfies: This ensures that the disc-shaped probe will not fall into the gap between the blades during the scanning process, and enables uninterrupted continuous scanning along the circumference of the blade tip.

[0071] 3) Equipment parameter configuration: Set the scanning path in the CMM control system (continuous scanning along the circumferential direction of the blade tip, supporting multi-height level scanning, scanning step size). (Set according to blade accuracy requirements, default 0.05-0.1mm), probe contact force (0.1-0.3N, to avoid damaging the blade tip surface).

[0072] 4) Input of core parameters: Input the key structural parameters of the disc-shaped probe (diameter) ,thickness (Unit: mm) Blade assembly parameters (number of blades) Installation angle (Unit: °), Measurement configuration parameters (number of scan layers) Scanning height of blade cross section at each level The data is entered into the analysis system, providing a foundation for subsequent data processing, error compensation, number matching, and leaf tip segmentation.

[0073] 2. Calibration of the first blade contact point:

[0074] 1) Calibration basis: Based on the design drawings of the engine blade to be measured, determine the target calibration area of ​​the blade tip of blade No. 1 (usually the middle height position of the blade tip).

[0075] 2) Calibration Operation: Using the CMM-driven disc probe, precisely touch the calibration area at the tip of the first blade (the middle height position of the blade tip in each level), and record the three-dimensional coordinates of this touch point. The corresponding polar angle in the two-dimensional polar coordinate system is θ. l0 .

[0076] 3) Calibration Purpose: This touch point serves as the "blade reference for the current level," ensuring that the blade data scanned at each level matches the actual blade number. , (The number of leaves corresponds one-to-one).

[0077] 3. Continuous scanning sampling of leaf tip curves:

[0078] 1) The CMM-driven disc probe performs adaptive circular continuous scanning along the annular contour of the entire blade tip of the aero-engine, supporting multi-height and multi-level scanning along the axial direction; due to the diameter of the disc probe... Greater than the maximum gap between the tips of adjacent leaves The probe remains in constant contact with the curved surface of the blade tip, without jamming or detachment, ensuring continuous and uninterrupted sampling at each level.

[0079] 2) Data Acquisition: Synchronously acquire the three-dimensional coordinate data of the blade tip surface at each height level ( (Unit: mm) and normal vector data of probe contact point .

[0080] like Figure 4 , Figure 5 The image shows scan data of a stator blade at a certain height in a single-layer casing.

[0081] Sampling advantages: ① Compared with single-point sampling, continuous scanning can obtain complete blade tip curve data, avoid missing extreme points, and ensure accurate blade tip radius and good repeatability; ② Based on the selection principle of "diameter greater than maximum gap", the scanning process is fast and reliable.

[0082] 3) Data Output: After scanning each height level, the output includes continuously sampled points (≥1000 points). Point cloud dataset And associate it with the corresponding scan layer height. (Unit: mm), used as raw data for subsequent analysis.

[0083] 4. Rotor / Stator Blade Tip Type Determination:

[0084] For each level of continuously scanned point cloud data set, determine the vector direction. If the vector direction points to the center of the circle, it is the stator blade scan data; otherwise, if the vector direction is away from the center of the circle, it is the rotor blade scan data.

[0085] 5. Adaptive search segmentation algorithm based on angle interval:

[0086] High-density continuous data sets acquired by CMM disc probe scanning at each level (The spacing between adjacent scanning points in the leaf tip region is uniform.) Based on the geometric characteristics of the leaf's annular distribution, a technical approach of "theoretical angle interval guidance + angle sector extreme value search + adaptive iterative positioning" is adopted to achieve automated and robust segmentation of leaf tip scanning data. Taking the stator leaf scanning data as an example, the specific steps are as follows:

[0087] 5.1 Data Preparation and Polar Coordinate Transformation:

[0088] 1) Input the circumferential scan point cloud data set, i.e., the continuous scan point cloud data set of the leaf tip. .

[0089] 2) Polar coordinate transformation: Establish a polar coordinate system with the engine axis as the origin (the axis coordinates are pre-calibrated using CMM), and transform the two-dimensional coordinates in the rectangular coordinate system into polar coordinates. Convert to polar coordinate parameters:

[0090] a) Polar radius calculation: , where R is the radial distance from the sampling point to the engine axis (unit: mm).

[0091] b) Polar angle calculation: , where θ is the circumferential angle of the sampling point (unit: °).

[0092] 3) Data association and storage: This involves storing the converted polar coordinate data. Compared with the original input data By linking them one by one, a comprehensive data set is formed.

[0093] Determine the segmented polar angle based on the polar radius R and the polar angle θ. ,like Figure 6 As shown.

[0094] 5.2 The first segment polar angle ( )Sure:

[0095] 1) Search for a circular scan point cloud dataset The sampling point with the largest radius in the middle is denoted as Its polar angle is .

[0096] 2) Extraction In the circumferential scan point cloud data set Pre-order neighbor in the sequence Adjacent points to the next order The corresponding polar angles are respectively and .

[0097] 3) Calculate the absolute difference of polar angles:

[0098] .

[0099] 4) Select the neighboring points corresponding to the larger difference as... The corresponding polar angle is :

[0100] like ,but , ;

[0101] like ,but , .

[0102] 5) Calculation and The average value of the polar angles yields the first segmented polar angle. The formula is:

[0103] .

[0104] 5.3 Remaining segmented polar angles ( Iterative calculation and circular scanning point cloud dataset Segmentation:

[0105] 1) Iterative initialization:

[0106] a) Set the first segment polar angle .

[0107] b) The total number of blades is The theoretical angular interval corresponding to a single blade is .

[0108] c) Define the circular scan point cloud data set The data set in polar coordinates is ,in To the total number of scan points, For the first The polar radius of each point Normalized polar angle ( ).

[0109] Objective: To convert the dataset according to Divided into non-overlapping and full coverage Individual Cloud Subset .

[0110] 2) Iterative segmentation based on the number of blades ( ):

[0111] a) Segmented splitting angle update: , obtained the The theoretical perspective of segmenting each segment.

[0112] b) Definition of angle sector: (in...) Centered on, delineate the first Each segment of angle sector .

[0113] c) Sector normalization processing: If Then the sector is split into ;like Then the sector is split into .

[0114] d) Determine the point with the maximum polar radius within the sector. In the sector (or The point corresponding to the maximum polar radius within the search area. ,

[0115] The formula is: .

[0116] e) Extraction Pre-order neighbor points in the scan sequence Adjacent points to the next order The corresponding polar angles are respectively and .

[0117] Calculate the absolute difference of polar angles:

[0118] ;

[0119] Select the neighboring points corresponding to the larger difference as :

[0120] like ,but , ;

[0121] like ,but , .

[0122] f) Determine the segmented polar angle :

[0123] calculate and The average value of the polar angles yields the first... Segmented polar angles The formula is: .

[0124] g) Segmentation of circular scan point cloud data: (The first...) The point cloud subsets corresponding to each segment are: (Corresponding to sector splitting) ).

[0125] Integrity check: All segments satisfy the requirements. This enables non-overlapping, full-coverage segmentation of circumferential data.

[0126] 3) Segmented polar angle sequence calibration:

[0127] For all segmented polar angles Perform normalization calibration to ensure all fall within the range Within the range:

[0128] like ,but ;

[0129] like ,but .

[0130] Output: Includes A sequence of segmented polar angles and the corresponding A set of points on a circle segmented into cloud subsets The segments are divided according to the number of blades.

[0131] The outlier removal method, based on iterative residual statistical analysis and adaptive spline fitting, is as follows:

[0132] 3.1) Data preprocessing and segmented modeling:

[0133] A set of points on a circle segmented into cloud subsets Polar coordinates, the corresponding set of rectangular coordinate systems ,in, The total number of segments. The subset of the rectangular coordinate system corresponding to the i-th circular segment point cloud subset Si. ,in:

[0134] This represents the total number of initial sampling points for the circular segment;

[0135] Indicates the first Cartesian coordinates of each sampling point;

[0136] The incrementing order of index i strictly corresponds to the physical scan trajectory of CMM, ensuring that adjacent index points have geometric continuity and topological adjacency.

[0137] 3.2) Iterative outlier detection and filtering algorithm:

[0138] A closed-loop iterative mechanism is employed, gradually purifying the dataset through a cyclical process of "fitting - residual evaluation - outlier removal." The first... The effective dataset after the next iteration is ,in, For the first The number of remaining valid points in the next iteration, and the initial state. This is the complete original data.

[0139] 3.2.1) Adaptive smooth spline fitting:

[0140] In the In this iteration, a cubic univariate spline function is constructed. To approximate the current dataset The geometric shape. The fitting process aims to minimize the objective functional, which includes a data fidelity term and a smoothness penalty term. :

[0141]

[0142] in:

[0143] This represents the space of functions whose second derivatives are continuous within their domain.

[0144] ;

[0145] First item The sum of squared residuals ensures that the fitted curve closely approximates the observed data;

[0146] Second item This is a curvature integral penalty term used to suppress non-physical oscillations caused by overfitting;

[0147] This is a smoothing factor used to adjust the balance between the fitted curve's sensitivity to noise and its smoothness (in some embodiments) ).

[0148] 3.2.2) Residual sequence calculation:

[0149] Based on fitting function Calculate the vertical residual for each sampling point in the current dataset and construct the residual sequence. :

[0150]

[0151] The residual sequence characterizes the degree of deviation of the observed points from the theoretical smoothing trend.

[0152] 3.2.3) Double Robust Statistical Criterion:

[0153] To overcome the limitations of a single statistic in complex noise environments, this embodiment proposes a method that fuses the median absolute deviation (MAD) and standard deviation (MAD). A dual discrimination mechanism is used to construct a binary anomaly mask.

[0154] (a) Robustness assessment based on MAD:

[0155] Calculate the median of the residual sequence and median absolute deviation :

[0156] ;

[0157] ;

[0158] Set dynamic threshold (Preferred) Generate the first type of exception marker. :

[0159]

[0160] in This is an indicator function, taking the value 1 if the condition is met, and 0 otherwise. This criterion is highly robust to non-Gaussian distributed heavy-tailed noise and extreme outliers.

[0161] (b) Based on Normal distribution discrimination:

[0162] Calculate the standard deviation of the residual series :

[0163]

[0164] Set dynamic threshold (Preferred) Generate a second type of exception marker. :

[0165]

[0166] This criterion is based on the normal distribution assumption and can effectively identify significant deviations from the Gaussian noise model.

[0167] (c) Generation of comprehensive anomaly mask:

[0168] The two discrimination results are combined using a logical "OR" operation to maximize the detection rate of outliers: If there exists any j such that If so, it is determined that there are outliers in the current iteration batch that need to be removed.

[0169] 3.3) Iterative convergence and data reconstruction mechanism

[0170] 3.3.1) Dataset update strategy:

[0171] Define the set of outlier indices detected in the k-th iteration as follows: Perform data pruning to generate the input dataset for the next iteration. :

[0172]

[0173] At the same time, a global exception index mapping table is established. Record all removed points in the original dataset. The corresponding index in the database is used to achieve source tracing.

[0174] 3.3.2) Termination condition determination:

[0175] The algorithm executes a loop iteration and stops when any of the following termination conditions are met:

[0176] Convergence criterion: This means that no new anomalies were detected in the current round, and the data distribution has stabilized.

[0177] Iteration limit: number of iterations (To prevent infinite loops, the best option is...) );

[0178] Sample protection criterion: number of remaining valid points (Preferred) This is to avoid the spline fitting matrix becoming singular or failing due to insufficient samples.

[0179] 3.3.3) Final Data Reconstruction:

[0180] After the iteration terminates, a clean dataset is obtained. .

[0181] By introducing a dual statistical discrimination mechanism, the problem of high false negative or false positive rates in the traditional single threshold method when facing mixed noise (Gaussian noise superimposed with impulse noise) is effectively solved. At the same time, the iterative adaptive fitting strategy can dynamically adjust the baseline curve, avoid the distortion effect of the initial outlier on the fitting model, and significantly improve the robustness and reconstruction accuracy of the circumferential contour data processing.

[0182] The outlier detection and removal method in this embodiment has the following advantages:

[0183] a. Purely data-driven, without relying on theoretical models:

[0184] The entire process is based on the actual shape of the point cloud for adaptive spline fitting, without relying on standard numerical models, theoretical contours or preset theoretical values, and is perfectly adapted to detection scenarios without theoretical benchmarks.

[0185] b. Excellent resistance to mixed noise:

[0186] By integrating MAD and 3σ dual statistical discrimination, it can simultaneously handle Gaussian noise, impulse outliers, and heavy-tailed noise. Compared with single discrimination methods, it has a lower rate of false negatives and false positives.

[0187] c. Iterative fitting and self-correction to avoid abnormal interference:

[0188] A closed-loop iteration of "fitting-identification-removal" is adopted to dynamically update the fitting benchmark, avoid the distortion of the overall contour trend by initial outliers, and ensure that the fitting curve fits the actual measured blade tip contour.

[0189] d. The smoothing constraints are reasonable, restoring the true shape:

[0190] Spline fitting introduces a curvature penalty term to suppress overfitting and spurious oscillations, accurately restoring the geometric shape of the circular contour itself.

[0191] e. The operating mechanism is safe and stable:

[0192] The triple iteration termination condition prevents infinite loops and insufficient samples from causing fitting failure. The algorithm is robust and adaptable to engineering test data.

[0193] 6. The tip radius value is matched with the blade number (based on blade number 1 as the baseline):

[0194] Blade number matching: Use blade number one to calibrate the reference polar angle θ l0For reference, calculate the polar angle of each finely positioned blade tip vertex. With θ l0 The difference △θ=∣θ i -θ l0 |; The segment corresponding to the leaf tip with the smallest difference is leaf number 1. The remaining segments are numbered sequentially from 2 to N in a clockwise or counterclockwise direction to ensure that all leaf tip data at this level are accurately associated with the actual leaf number.

[0195] Consistency of tip apex numbering across different heights within the same level: The tip apex numbering calculated for other heights within the same level is based on the θ value calibrated at the contact point of the first blade in that level. l0 Matching is performed based on a benchmark to ensure that the tip apex number of the same blade at different heights is consistent and traceable.

[0196] 7. Stylus cosine error compensation:

[0197] Considering the structural characteristics of the CMM paired with the probe, an error compensation model is designed to couple the probe parameters with the blade mounting angle, thus addressing measurement deviations caused by the probe's own structure and the blade mounting attitude.

[0198] like Figure 7 As shown, the error source analysis is as follows: the blade installation angle is... When the spherical area of ​​the disc-shaped probe contacts the curved surface of the blade tip, the height of the probe center is inconsistent with the height of the contact point, resulting in a shift between the actual contact point of the probe and the theoretical blade tip apex, thus forming a cosine error.

[0199] Compensation amount R is the radius of the probe's arc surface, compensated along the normal direction of the contact point (I / J / K vector direction).

[0200] Coordinate compensation: Based on the calculated ΔD, the blade tip coordinates are corrected in the reverse direction along the contact point normal.

[0201] ;

[0202] .

[0203] in, , To compensate for the rectangular coordinates of the tip of the blade (unit: mm).

[0204] 8. Calculation of ultimate radius:

[0205] Based on the compensated coordinates Calculate the distance from the tip of the blade to the engine axis. (Unit: mm), as the final polar radius of the blade tip.

[0206] The core innovations of this embodiment lie in the measurement scheme of "CMM + large-diameter probe", the adaptation principle of "probe diameter greater than the maximum gap between blades", the numbering and traceability mechanism for the calibration of blade number one, the deep integration of the adaptive search segmentation algorithm based on angle interval and the special error compensation model, which has the following beneficial effects:

[0207] 1. Significantly improved measurement speed and reliability: clearly defines the diameter of the disc-shaped probe. Greater than the maximum gap between the tips of adjacent leaves The selection principle fundamentally avoids scanning interruptions, measurement failures, or equipment damage caused by the probe falling into the gap, ensuring the smooth completion of continuous circumferential scanning. Compared with touch point measurement, scanning measurement is 10-20 times more efficient.

[0208] 2. Optimization of sampling accuracy and stability: The disc structure of the disc probe has a larger contact area with the blade tip curved surface than the spherical probe, resulting in higher sampling stability and stronger anti-noise capability; the CMM continuous scanning method acquires high-density, continuous blade tip curve data, avoiding the problem of missing extreme points in single-point sampling, and providing a high-quality data foundation for subsequent accurate segmentation and blade tip point positioning.

[0209] 3. Improved Segmentation Accuracy and Robustness: The automatic blade tip segmentation algorithm based on angle intervals utilizes the geometric characteristics of the uniformly distributed ring of the blade. Guided by theoretical angle intervals, it employs angle sector extreme value search and adaptive iterative positioning. It does not rely on point distance statistics and is unaffected by measurement noise and sampling density fluctuations, thus tolerating [variable issues]. Within the acceptable range of blade installation errors and machining deviations, the segmentation accuracy reaches 100%, effectively solving the problems of poor adaptability and weak anti-interference ability of traditional segmentation methods.

[0210] 4. Improved error compensation: Addressing the thickness of the probe. ,diameter Features and installation angle A special error compensation model was designed to solve the coupling error between the probe structure and the installation posture.

[0211] 5. Strong traceability through numbering: The reference polar angle is obtained by calibrating the test point of blade number one. This enables precise matching of blade numbers at different height levels, ensuring that measurement data for the same blade is traceable across the entire height range.

[0212] 6. High degree of automation: It realizes end-to-end automation of "probe adaptation verification → blade calibration → CMM scanning → automatic data import → automatic blade tip segmentation → data processing → error compensation → analysis → report", without the need for manual intervention in segmentation, point sampling and probe selection verification. The measurement efficiency is 10-20 times higher than traditional manual measurement, which is suitable for batch production testing needs.

[0213] 7. High adaptability: Through parameter configuration ( , , , (etc.), can be adapted to the tip measurement of blades / rotors of different types of aero-engines, and is compatible with both inner and outer tip types. The adaptive search segmentation algorithm based on angle intervals can adapt to the segmentation requirements of different numbers of blades, and has a wide range of applications.

[0214] 8. High reliability of results: The precise positioning and outlier data removal of the segmented algorithm effectively offset measurement noise and system errors, meeting the high-precision testing requirements of aero-engines.

[0215] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions claimed by the present invention.

Claims

1. A method for measuring the tip radius of a rotor-stator blade in an aero-engine, characterized in that, The method includes: A probe was used to perform layered scanning of the blade tips at different heights perpendicular to the engine axis, obtaining three-dimensional coordinate data (X, Y, Z) of the blade tip at each height, and normal vector data (I, J, K) of the contact point; the diameter of the probe... , This is the maximum gap between the tips of adjacent leaf blades; In each scanning height plane, a polar coordinate system is established with the engine axis as the origin, and the planar coordinates in the three-dimensional coordinate system are taken. Convert to polar coordinate data (polar radius R, polar angle θ); The segmented polar angle is determined based on the polar radius R and the polar angle θ. The polar coordinate data is then segmented based on the segmented polar angle, and each segmentation results form a subset of point cloud. For each subset of point clouds, determine the blade tip vertex and calculate the distance from each blade tip vertex to the engine axis as the blade tip polar radius. Methods for determining the segmented polar angle based on the polar radius R and polar angle θ include: Determine the sampling point with the largest polar radius The corresponding polar angle is ; extract Preorder neighbor Adjacent points to the next order The corresponding polar angles are respectively and ; calculate and , The absolute difference is used to select the neighboring points corresponding to the larger difference as... The corresponding polar angle is ; calculate and The average value of the polar angles yields the first segmented polar angle. ; The theoretical angular interval corresponding to a single blade is , Total number of blades; Segmentation and splitting angle update, , obtained the Theoretical search perspectives for each segment; Determine the angle sector, in order to Centered on, delineate the first Each segment of angle sector ; In sector The point corresponding to the maximum value of the inner search polar radius ; extract Preorder neighbor Adjacent points to the next order The corresponding polar angles are respectively and ; calculate and The absolute difference is used to select the neighboring points corresponding to the larger difference as... The corresponding polar angle is ; calculate and The average value of the polar angles is used to obtain the i-th segment polar angle. ; The coordinate data is segmented based on the segmented polar angle.

2. The method for measuring the tip radius of aero-engine stator blades according to claim 1, characterized in that, After determining the angle sector, the sector is normalized. Then the sector is split into ;like Then the sector is split into ; For in sectors or The point corresponding to the maximum value of the inner search polar radius .

3. The method for measuring the tip radius of the rotor-stator blades of an aero-engine according to claim 1, characterized in that, For all segmented polar angles Normalization calibration is performed to form a normalized and calibrated segmented point cloud subset. For each segmented point cloud subset, outliers are detected and removed based on iterative residual statistical analysis and adaptive spline fitting, using a dual robust statistical discrimination criterion that integrates the absolute deviation of the median and the standard deviation.

4. The method for measuring the tip radius of aero-engine rotor-stator blades according to claim 1, characterized in that, Blade installation angle is ; Compensation amount R is the radius of the probe's arc surface; Correcting the leaf tip vertex coordinates: ； ； By calculating the distance from the blade tip to the engine axis .

5. The method for measuring the tip radius of aero-engine stator blades according to claim 1, characterized in that, The probe touches the calibration area at the tip of blade number one, and the three-dimensional coordinates of that touch point are recorded. The corresponding polar angle is θ l0 .

6. The method for measuring the tip radius of aero-engine rotor-stator blades according to claim 5, characterized in that, Calculate the polar angle at each leaf tip vertex. With θ l0 The difference is calculated, and the segment corresponding to the tip of the leaf with the smallest difference is leaf number 1. The remaining segments are numbered sequentially from 2 to N in a clockwise or counterclockwise direction.

7. The method for measuring the tip radius of aero-engine rotor-stator blades according to claim 6, characterized in that, The blade tip apex numbering for other heights in the same level is based on the polar angle θ calibrated by the contact point of the first blade in that level. l0 Matching is performed based on the baseline.

8. A system for measuring the tip radius of aero-engine stator blades, characterized in that, The measurement system includes a probe and a data processing system, wherein the data processing system performs the measurement according to the measurement method described in any one of claims 1-7.

Images