Method and apparatus for detecting delamination defects in flexible microwave resonators made of ceramic matrix composites

Abstract

This invention discloses a method and apparatus for detecting delamination defects in ceramic matrix composites using a flexible microwave resonator. The apparatus includes a flexible microwave resonator, a vector network analyzer, a two-dimensional scanning platform, and a data processing unit. The flexible microwave resonator consists of a polyimide substrate, a top-opening resonant ring, and a bottom coplanar waveguide feeding structure, conformally attached to the surface to be tested due to its flexibility. The method acquires S11 curves at various points by scanning, extracts the resonant frequency as a feature value, and finds that the resonant frequency offset has a monotonically changing relationship with the depth of the delamination defect. By constructing a resonant frequency distribution image, high-sensitivity localization, imaging, and quantitative depth inversion of delamination defects in ceramic matrix composites are achieved. This invention solves the problems of rigid sensors being unable to adapt to curved surface detection and low sensitivity for small defects, and is suitable for non-destructive testing of complex components such as aero-engine blades.

Description

Technical Field

[0001] This invention belongs to the field of nondestructive testing technology, specifically relating to a method and device for detecting delamination defects in ceramic matrix composites based on flexible microwave resonators. Background Technology

[0002] Ceramic matrix composites (CMCs) are widely used in aero-engine hot-end components, aerospace thermal protection structures, and nuclear energy fields due to their excellent properties such as lightweight, resistance to extreme high temperatures, and non-brittle fracture. However, in the manufacturing process of SiC fiber-reinforced ceramic matrix composites (such as 2D SiC / SiC), delamination is a common and extremely harmful internal defect that seriously affects the mechanical properties and service life of the components.

[0003] Existing non-destructive testing methods, such as X-ray inspection and infrared thermography, have many limitations when applied to ceramic matrix composites: X-ray inspection is insensitive to delamination defects parallel to the X-ray direction and poses radiation safety hazards; the equipment is expensive and the inspection efficiency is low; although infrared thermography has a faster inspection speed, its detection depth is shallow (usually ≤1mm), making it difficult to detect deep defects, and its spatial resolution for small defects is limited (usually around 1mm, comparable to or slightly lower than this method, but far inferior to this method in terms of quantitative depth inversion capability). Near-field microwave non-destructive testing technology, with its advantages of non-contact operation, high resolution, strong penetration, and no ionizing radiation, has become an effective way to solve the above problems. In existing technologies, such as the patent CN117169340A which discloses a detection method based on a rigid planar resonator, the sensor is a rigid structure and must be kept perpendicular to the surface being tested, making it unsuitable for complex curved surfaces such as aircraft blades; furthermore, its detection effect relies on reflection from a metal backplate, and its sensitivity decreases significantly on ceramic matrix composites without a metal backplate.

[0004] Therefore, there is an urgent need to develop a detection method and device that can conformally fit onto curved surfaces, has high sensitivity to delamination defects in ceramic matrix composites, and does not require the assistance of a metal backing plate. Summary of the Invention

[0005] The purpose of this invention is to provide a method and device for detecting delamination defects in ceramic matrix composites based on flexible microwave resonators, so as to solve the problems in the prior art that rigid sensors cannot adapt to curved surface detection, have insufficient sensitivity to near-surface micro-delamination defects, and are difficult to achieve quantitative characterization of defect burial depth.

[0006] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, the present invention provides a flexible microwave resonator for detecting delamination defects in ceramic matrix composite materials, comprising: A flexible microwave resonator includes a flexible dielectric substrate, an open resonant ring fixed on the top of the flexible dielectric substrate, and a coplanar waveguide feeding structure fixed on the bottom of the flexible dielectric substrate. A vector network analyzer is used to transmit microwave signals to the flexible microwave resonator and acquire the S11 curve. A coaxial transmission line connects the flexible microwave resonator to the vector network analyzer; A two-dimensional scanning platform is used to control the flexible microwave resonator to perform two-dimensional scanning along the surface of the ceramic matrix composite material to be tested. The data processing unit is used to construct a scanned image of the material under test based on the resonant frequency of the S11 curve at each scanning point, so as to realize the detection of layered defects.

[0007] Furthermore, the coplanar waveguide feeding structure includes a signal transmission line and grounded copper foils located on both sides of the signal transmission line. A gap is provided between the signal transmission line and the grounded copper foils on both sides, and an SMA connector for connecting a vector network analyzer is provided at one end of the signal transmission line. This structural design ensures efficient transmission and coupling of microwave signals, reduces reflection loss, and improves detection sensitivity. By precisely controlling the gap width, precise matching of characteristic impedances (typically 50Ω) can be achieved, thereby minimizing signal reflection and improving the Q value of the resonant peak.

[0008] Furthermore, the signal transmission line faces and crosses the central region of the split-ring resonator, exciting its resonant response through near-field electromagnetic coupling. This coupling method eliminates the need for physical contact, avoiding mechanical wear while ensuring signal stability and repeatability. The strength of the near-field coupling is closely related to the relative position between the signal transmission line and the split-ring resonator. This invention, through optimized design, achieves an optimal coupling coefficient, ensuring that the resonator generates a significant resonance peak within the operating frequency band.

[0009] Furthermore, the flexible dielectric substrate is made of polyimide thin-film circuit board, with a preferred thickness of 0.03mm to 0.1mm. Polyimide material has excellent flexibility, high temperature resistance (long-term operating temperature can reach above 250℃), and low dielectric loss characteristics, making it suitable for high-frequency microwave detection of complex curved surface components. The selection of substrate thickness must balance mechanical flexibility and electromagnetic performance: too thin may lead to insufficient mechanical strength, while too thick will reduce flexibility and increase dielectric loss.

[0010] Furthermore, the split-ring resonator is a copper foil ring with a feeding notch. This structure exhibits a high quality factor (Q value typically reaching 100~300) in the microwave band and is extremely sensitive to changes in dielectric constant, effectively capturing local dielectric disturbances caused by delamination defects. The resonant frequency of the split-ring resonator is determined by the ring's equivalent inductance and capacitance. When delamination defects exist within the material under test, changes in the local dielectric constant alter the ring's equivalent capacitance, resulting in a shift in the resonant frequency.

[0011] Furthermore, the gap width between the signal transmission line and the grounded copper foil on both sides is 0.05mm~0.2mm; the outer diameter of the open resonant ring is 3mm~10mm, the line width is 0.2mm~0.6mm, and the width of the feed notch is 0.05mm~0.2mm. These dimensional parameters can be optimized and adjusted according to the dielectric properties of the material under test and the defect detection depth to achieve the best detection resolution and sensitivity. For example, for materials with high dielectric constants, the size of the resonant ring can be appropriately reduced to improve spatial resolution; for deep defect detection, the size of the resonant ring can be increased to enhance penetration depth.

[0012] Secondly, the present invention provides a method for detecting delamination defects in ceramic matrix composite materials based on a flexible microwave resonator, applied to the detection device described in any of the above claims, comprising the following steps: Step 1: Apply the flexible microwave resonator with an open resonant ring to the surface of the ceramic matrix composite material to be tested; Step 2: Emitter a microwave signal into the flexible microwave resonator using a vector network analyzer and acquire the S11 curve at the current detection point; Step 3: Extract the resonant frequency from the S11 curve as the microwave signal feature value of the current detection point; Step 4: Control the flexible microwave resonator to perform a two-dimensional scan along the surface of the ceramic matrix composite material to be tested, and repeat steps 2 to 3 to obtain the resonant frequency corresponding to each scanning point; Step 5: Using the coordinates of the scanning points as the position index and the offset of the resonant frequency of the S11 curve as the pixel value, construct a scanned image of the ceramic matrix composite material to be tested, and identify the delamination defect area through image processing algorithms, and invert the defect burial depth based on the frequency offset.

[0013] Furthermore, in step three, if layered defects exist within the material under test, the extracted resonant frequency will shift, and this shift monotonically changes with the defect depth. This principle provides the theoretical basis for the quantitative inversion of defect depth in this invention. By establishing a finite element simulation model or experimental calibration, the quantitative relationship between the frequency shift and the defect depth can be obtained, typically exhibiting an exponential decay or polynomial fitting function.

[0014] Furthermore, in step one, the flexible microwave resonator is conformally applied to the surface of the ceramic matrix composite material under test, which has a curved structure, utilizing its flexibility. This characteristic enables the present invention to be applied to the non-destructive testing of complex curved components such as aero-engine blades and combustion chamber liners. In practical applications, a tight fit between the resonator and the curved surface can be ensured through vacuum adsorption or flexible tape, avoiding the influence of air gaps on the test results.

[0015] Furthermore, in step five, the data processing unit also performs the following operations: setting a threshold for the resonant frequency offset, and marking regions in the scanned image whose resonant frequency offset exceeds the threshold as layered defect regions; performing connected component analysis on the marked regions to extract the area and location of each defect region; and mapping the average offset of each defect region to the burial depth according to a pre-calibrated frequency offset-burial depth relationship curve. This process enables automatic defect identification and quantitative assessment, improving the level of intelligence in detection.

[0016] The beneficial effects of this invention are as follows: 1. High sensitivity and high precision: Employing an open resonant ring structure, it exhibits highly sensitive response to localized dielectric anomalies in ceramic matrix composites. Compared to traditional echo amplitude and phase detection, the resonant frequency change of the S11 curve is more significant, enabling precise capture of minute near-surface delamination defects.

[0017] 2. Excellent adaptability to curved surfaces: The flexible microwave resonator uses flexible substrates such as polyimide, which can be conformally applied to the surface of complex curved components such as aero-engine blades, solving the problem that rigid sensors cannot fit curved surfaces and ensuring the consistency and accuracy of detection.

[0018] 3. Quantitative characterization of defects: The monotonic relationship between resonant frequency offset and the burial depth of layered defects was discovered. This not only allows for the location of defects but also enables the inversion of the burial depth of defects through frequency offset, achieving a leap from qualitative detection to quantitative assessment.

[0019] 4. Non-destructive and safe: The entire process uses microwave non-contact detection, which has no ionizing radiation, requires no coupling agent, and will not cause any damage to the high-value ceramic matrix composite material being tested.

[0020] 5. High detection efficiency: Combining a two-dimensional automatic scanning platform with a fast signal processing algorithm, it can achieve large-area rapid scanning, meeting the needs of engineering sites for detection efficiency. Attached Figure Description

[0021] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1This is a schematic diagram of the structure of the flexible microwave resonator of the present invention; Figure 2 This is a schematic diagram illustrating the delamination defect detection principle of the ceramic matrix composite material of the present invention. Figure 3 This is a schematic diagram of a ceramic matrix composite specimen with delamination defects of different depths according to the present invention; Figure 4 This is a typical resonance response curve of the flexible microwave resonator of the present invention in a defect-free region (background region); Figure 5 This is a graph showing the resonant frequency shift curves corresponding to different depths of layered defects in this invention. Figure 6 This is a distribution diagram of the two-dimensional scanning results of the ceramic matrix composite material specimen of the present invention; Figure 7 This is a schematic diagram of the flexible microwave resonator conformally applied to the turbine blade of an aero-engine according to the present invention.

[0022] Numbers in the diagram: 1. Grounding copper foil; 2. Signal feeder (i.e., signal transmission line); 3. Flexible dielectric substrate; 4. Open resonant ring; 5. SMA connector; 6. Gap; 7. Feeding notch; 100. Flexible microwave resonator; 200. Aero-engine turbine blade. Detailed Implementation

[0023] 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.

[0024] Example 1: Construction of the detection device like Figure 1 As shown, this embodiment provides a flexible microwave resonator 100, the core structure of which includes: 1. Flexible dielectric substrate 3: Selected with dielectric constant ε r A polyimide film with a viscosity of 2.65 and a thickness of 0.03 mm exhibits excellent flexibility and high-temperature resistance. This material also demonstrates low-loss characteristics in the microwave frequency band, ensuring a high Q value for the resonant ring.

[0025] 2. Open-ended resonant ring 4: Located at the top of the substrate, it is made of copper foil with a thickness of 0.017 mm. It is designed as a circular ring with a feed notch 7, an outer diameter of 5 mm, a linewidth of 0.4 mm, and a feed notch width of 0.1 mm. This structure serves as a sensing element, and its resonant characteristics are highly sensitive to changes in the dielectric of the material under test. Through simulation optimization, this size was selected to obtain a significant resonance peak in the 23–25 GHz frequency band, facilitating signal acquisition and analysis.

[0026] 3. Coplanar Waveguide (CPW) Feed Structure: Located at the bottom of the substrate, it consists of a central signal feed line 2 (signal transmission line) and two ground copper foils 1 on either side. The signal feed line is 0.28 mm wide, and the gap 6 between it and the ground copper foil is 0.1 mm to ensure 50Ω characteristic impedance matching. The signal feed line crosses the center of the open resonant ring and is excited to resonance through near-field electromagnetic coupling. An SMA connector 5 is soldered to the end of the signal feed line for connecting a vector network analyzer.

[0027] The complete testing setup also includes: a vector network analyzer (VNA, model Keysight N5222B, frequency range 10MHz~26.5GHz, dynamic range >130dB), a low-loss coaxial cable (model SMA-SMA, length 1.5m, insertion loss <0.5dB@5GHz), a high-precision two-dimensional motorized scanning platform (travel 200mm×200mm, positioning accuracy ±0.01mm, repeatability ±0.005mm), and a host computer data processing unit (equipped with an Intel Core i7 processor, 16GB of memory, running a custom LabVIEW control program). The VNA is connected to the flexible microwave resonator 100 via the coaxial cable, responsible for emitting microwave signals (frequency band covering 10MHz-26.5GHz, output power +13dBm) and acquiring the reflection coefficient S11 curve. The scanning platform moves the resonator on the sample surface at a preset step size (e.g., 0.5mm), with an adjustable moving speed (typically 10mm / s). The data processing unit controls the scanning process, extracts the resonant frequency in real time, and generates a two-dimensional frequency distribution image.

[0028] To verify the detection performance of different size parameters, this embodiment also designed multiple sets of comparative experiments: Comparison of resonant ring outer diameters: Resonant rings with outer diameters of 3mm, 5mm, and 8mm were fabricated and scanned on specimens with the same defects. The results show that the resonant ring with an outer diameter of 5mm achieves the best balance between resolution and sensitivity, and can detect defects as small as 0.8mm × 0.8mm. While the resonant ring with an outer diameter of 3mm has higher resolution (detecting defects as small as 0.5mm × 0.5mm), its penetration depth is shallower (<1.5mm). The resonant ring with an outer diameter of 8mm can achieve a penetration depth of 3mm, but its sensitivity to small defects decreases.

[0029] Substrate thickness comparison: Polyimide substrates with thicknesses of 0.03mm, 0.05mm, and 0.1mm were used respectively. The experiment showed that the thinner the substrate, the better the flexibility and the better the bonding effect, but the mechanical strength decreased slightly; the 0.03mm thickness ensured sufficient flexibility while still meeting the mechanical strength requirements for repeated use.

[0030] Gap width optimization: Three gap widths of 0.05mm, 0.1mm, and 0.2mm were compared, and the characteristic impedance was measured using a time domain reflectometer (TDR). The impedance was closest to 50Ω (measured at 49.8Ω) with a gap of 0.1mm, the return loss S11 < -25dB, and the signal transmission efficiency was the highest.

[0031] Example 2: Verification and Simulation Analysis of Detection Principle A simulation model was built using CST Microwave Studio software. The object under test was a SiC / SiC ceramic matrix composite plate (20mm × 20mm × 3mm), with an internal pre-fabricated air layer of 2mm × 2mm × 0.3mm to simulate delamination defects. The material parameters were set as follows: dielectric constant ε r = 8.5, loss tangent tanδ = 0.05. The mesh is a hexahedral mesh with a minimum mesh size of 0.05 mm. The solver is a frequency domain solver with a frequency range of 23~25 GHz and a step size of 1 MHz.

[0032] like Figure 3 and Figure 4 As shown, three operating conditions were set: a defect-free control group, a defect burial depth group of 1 mm, and a defect burial depth group of 2 mm. Simulation results show that the resonant frequency in the defect-free case is... f 0 = 23.602 GHz; when a delamination defect with a burial depth of 1 mm exists, the resonant frequency shifts to f 1 = 23.800 GHz; when the burial depth increases to 2 mm, the resonant frequency further changes to f =23.750 GHz. It can be seen that the introduction of layered defects shifts the resonant frequency towards higher frequencies, and the shift shows a non-monotonic decreasing trend with increasing burial depth, which may be related to the spatial distribution of the electromagnetic field and the scattering effect at the defect edge.

[0033] Simulation results are as follows Figure 5 As shown. When there are no defects, the resonant frequency is... f 0 = 23.602 GHz, S11 amplitude is -28dB. When a delamination defect with a burial depth of 1mm exists, the introduction of an air layer at the defect reduces the equivalent dielectric constant, causing the resonant frequency to shift towards higher frequencies. f 1 = 23.800 GHz, frequency offset Δ f = f 1 f 0 = +198 MHz, at which point the amplitude of S11 becomes -26dB; when the burial depth increases to 2mm, the resonant frequency shifts to f 2 = 23.750 GHz, frequency offset Δ f = f 2 f 0 = +148 MHz, S11 amplitude is -25dB.

[0034] Data shows that the presence of layered defects leads to a significant increase in the resonant frequency. As the burial depth increases from 1 mm to 2 mm, the frequency shift exhibits a non-monotonic trend (increasing first and then decreasing), which may be related to the coupling efficiency and edge effects of the electromagnetic field at specific burial depths. Nevertheless, the significant shift in resonant frequency (>100 MHz) and the attenuation of the S11 amplitude still confirm the effectiveness of utilizing the absolute value of the frequency shift (|Δ) f The feasibility of using ∣) and amplitude variation as characteristic values ​​to characterize the presence and burial depth of defects. To further verify the robustness of the proposed method, defect sizes were varied: 1mm×1mm, 2mm×2mm, and 3mm×3mm were set, with a burial depth of 1mm for all sizes. The results show that the larger the defect area, the more significant the frequency shift: a 1mm×1mm defect shifted by 25MHz, a 2mm×2mm defect by 70MHz, and a 3mm×3mm defect by 115MHz. The shift is positively correlated with the defect volume, exhibiting an approximately linear relationship.

[0035] Defect shape variation: The defect shapes were set to square, circle, and rectangle (2mm×4mm) respectively, while keeping the area the same (8mm²). The results show that the shape has little effect on the frequency offset (offset difference <5%), indicating that the method is not sensitive to the defect shape and has good universality.

[0036] Multi-layer defect interference: Two layered defects are simultaneously placed at distances of 1 mm and 2 mm from the surface. Simulation shows that a double-peak phenomenon occurs at the resonant frequency. The shift of the main peak is mainly affected by the shallower defect, while the secondary peak corresponds to the deeper defect. A multi-peak identification algorithm can be used to distinguish and detect multi-layered defects.

[0037] Effect of material anisotropy: The dielectric constant of the material is set to anisotropic (ε). x ≈8.5, ε y ≈7.5, ε z (≈9.0), compared with the isotropic case, the difference in resonant frequency offset is <3%, proving that the method is also applicable to anisotropic materials.

[0038] Example 3: Physical object scanning detection experiment and image reconstruction SiC / SiC specimens containing delamination defects at a depth of 1 mm were fabricated, with overall dimensions of 50 mm × 50 mm × 4 mm. The specimens were prepared using a chemical vapor infiltration (CVI) process, with a fiber volume fraction of 40% and a porosity of approximately 10%. The defects were achieved by carbonizing a pre-placed polyimide film (0.3 mm thick) during sintering to form an air layer.

[0039] The flexible microwave resonator 100 was tightly attached to the surface of the specimen, and the two-dimensional scanning program was started. Given the specimen size and testing requirements, this experiment focused on scanning a 10mm × 10mm area (100 square millimeters) in the middle of the specimen, with a step size of 0.5mm, a sampling time of 0.2s per point, and a total scanning time of approximately 80 seconds. The vector network analyzer settings were: frequency band 18-26.5GHz, number of sampling points 1601, intermediate frequency bandwidth 100Hz, and step scanning mode.

[0040] like Figure 6 As shown in the generated resonant frequency distribution map, the background region has a uniform frequency (approximately 23.3 GHz, standard deviation ±3 MHz), while a distinct high-frequency bulge region appears at the corresponding defect location (the lowest point is approximately 23.8 GHz, offset 50 MHz), clearly outlining the defect's contour and area. A frequency gradient transition band exists at the edge of the defect region, approximately 1-2 pixels wide (0.5-1 mm), reflecting the gradual electromagnetic field change effect at the defect edge.

[0041] To further verify the reliability of the test, the same specimen was scanned repeatedly (10 times), and the standard deviation of the frequency at each point was calculated. The results showed that the frequency fluctuation in the background area was <±2MHz, and the frequency fluctuation in the defect area was <±5MHz, proving that the method has good repeatability and stability.

[0042] Comparative experiment: In order to verify the overall performance of this method, the same specimens were used for infrared thermal imaging detection and X-ray CT detection for comparison.

[0043] Infrared thermal imaging detection: An active infrared thermal wave detection system was used. Results showed that infrared thermal imaging can detect near-surface defects, but for defects buried deeper than 1 mm, signal attenuation is severe, and the signal-to-noise ratio is significantly reduced. Although its spatial resolution can reach about 1 mm, comparable to this method, it cannot accurately invert the defect burial depth through frequency offset like this method, and it has a high rate of missed detection for deep defects.

[0044] X-ray CT inspection: A microfocus X-ray CT system was used. CT inspection can clearly present the three-dimensional morphology of internal defects with high spatial resolution (up to 0.1 mm), and can detect all defects in this experiment. However, CT inspection poses safety hazards related to ionizing radiation, has high equipment costs, and a long inspection time (approximately 1800 seconds per item), making it unsuitable for rapid on-site screening. Furthermore, for thin-layered defects parallel to the X-ray direction, the contrast of CT is sometimes insufficient.

[0045] This method has a signal-to-noise ratio of up to 25dB, requires no coupling agent, has no radiation hazard, and has high detection efficiency (approximately 80s per region). While maintaining high spatial resolution (0.5mm), it also possesses a unique ability to quantitatively invert the burial depth of defects, and its overall performance is superior to the aforementioned comparative methods.

[0046] Example 4: Application of curved surface component inspection like Figure 7 As shown, for a typical curved surface component such as an aero-engine turbine blade 200, the blade material is SiC / SiC, and the radius of curvature of the blade surface varies from 15mm to 50mm. Utilizing the flexibility of the flexible microwave resonator 100, it is conformally applied to the curved surface of the blade's blade base and fixed by vacuum adsorption to ensure a tight fit.

[0047] The scanning was performed while maintaining good adhesion, with a scanning area of ​​30mm × 30mm and a step size of 1mm, covering the central area of ​​the blade's leaf base. During the scanning process, the adhesion between the resonator and the blade surface was monitored in real time, and an image recognition algorithm was used to determine if any localized warping was present.

[0048] To further verify the adaptability of curved surfaces, comparative experiments were conducted using arc-shaped specimens with curvature radii of 50 mm, 30 mm, and 15 mm. Each specimen was pre-embedded with a delamination defect of the same size (2 mm × 2 mm × 0.3 mm) and the same embedding depth (1 mm).

[0049] The radius of curvature is 50 mm, and the frequency offset is 70.2 MHz, which is basically consistent with the flat plate specimen (70.0 MHz). Curvature radius 30mm: frequency offset 69.5MHz, error <1%; 15mm radius of curvature: frequency offset 68.8MHz, error <2%; A curvature radius of 10mm results in a frequency offset of 66.5MHz, an error of approximately 5%, and increases the difficulty of fitting, posing a risk of localized lifting.

[0050] Therefore, this device has excellent detection accuracy when the radius of curvature is ≥15mm, and is suitable for the curvature surfaces of most aero-engine blades.

[0051] Data processing and defect identification logic (detailed algorithm description) In step five, the data processing unit executes the following image processing and depth inversion algorithms, specifically implemented as follows: 1. Data Acquisition and Preprocessing: The scanning platform moves along a preset path, and at each scanning point (x... i , y j At point ), the vector network analyzer acquires the S11 curve, obtaining the amplitude data (complex form, from which amplitude and phase can be extracted) corresponding to discrete frequency points. The data processing unit smooths and filters the raw S11 data using a Savitzky-Golay filter (window length 11, polynomial order 3), reducing random noise while preserving the details of the resonance peaks.

[0052] Optionally, the amplitude data can be normalized to eliminate systematic biases between different scan points (e.g., using a moving average or Savitzky-Golay filter) to reduce noise.

[0053] 2. Resonant Frequency Extraction: For each smoothed S11 curve, the resonant frequency f(x,y) is extracted using a peak search algorithm. The specific method is as follows: Within a preset resonant frequency band (e.g., 23-25GHz), find the point where the amplitude of S11 is at its minimum, i.e., the point where |S11| is at its maximum; specific algorithm: Calculate the first-order difference to identify the point where the amplitude increases after decreasing; The candidate minimum points are sorted, and the point with the smallest amplitude is selected as the initial resonance peak. If multiple local minima exist (which may correspond to multimodal resonance or multilayer defects), record all minimum points and their amplitudes for subsequent analysis.

[0054] The accuracy is improved by using parabolic interpolation: take the resonance peak and its left and right neighboring points, fit a quadratic function, and take the derivative to obtain the peak frequency. The accuracy can reach the sub-MHz level (measured standard deviation <0.5MHz).

[0055] For multi-layer defects, a multi-peak identification algorithm is used: set an amplitude threshold (e.g., -15dB), retain all minimum points below the threshold, and sort them by frequency to correspond to defects of different depths.

[0056] 3. Determination of reference frequency: Method 1 (Local Reference): Select 5×5 scanning points in a defect-free area (such as the edge of the specimen) and calculate the average value of its resonant frequencies as the reference frequency f. ref .

[0057] Method 2 (Global Reference): Draw a frequency histogram for the entire graph and take the peak value (mode) as f. ref This is applicable to situations where there are unknown, defect-free areas.

[0058] Method 3 (Dynamic Reference): For large components, the sliding window method can be used, with the local mode of the frequency within the window as the benchmark to eliminate the influence of material inhomogeneity.

[0059] 4. Offset Calculation: Calculate the frequency offset Δf(x,y) = f for each scan point. ref - f(x,y). A positive offset indicates a decrease in frequency, usually corresponding to the presence of defects. For the case of multi-layer defects, the offset Δf of each resonance peak can be calculated. k ~(x,y), k=1,2,..., corresponding to different depths.

[0060] 5. Defect area identification: Set threshold T h T h The method for determining it is as follows: measure the standard deviation σ of the offset in the defect-free area, and take T. h = k·σ, where k is typically taken as 3~5 (corresponding to the 3σ principle). Typically, σ≈2MHz, T h Select 6-10MHz.

[0061] Generate a binary image B(x,y), if Δf(x,y)>T h If the result is positive, then B(x,y) = 1 (candidate defect point); otherwise, B(x,y) = 0. Perform connected component analysis on the binary image (e.g., using the 8-neighbor labeling algorithm) to aggregate adjacent defect candidate points into independent defect regions.

[0062] Remove isolated noise points with an area smaller than the preset minimum defect size (e.g., 0.5 mm²) to avoid false detections.

[0063] 6. Defect Parameter Calculation: For each connected region, calculate its area (number of pixels × squared step size), center coordinates, maximum offset, and average offset. Optional calculations of the defect region's shape factors (such as roundness and aspect ratio) can aid in determining the defect type.

[0064] 7. Deep Inversion: The relationship curve between frequency offset and defect burial depth is pre-calibrated through simulation or experiment. For example, a series of standard specimens with delamination defects at different burial depths are made, and their corresponding Δf is measured. The function d = g(Δf) is then fitted (usually a monotonically decreasing function, which can be fitted using a polynomial or exponential function). For each defect region, based on its average offset Δf avg Substituting the values ​​into the calibration curve, we obtain the burial depth d of the defect; Output an inspection report, which includes information such as the location, area, and depth of defects.

[0065] Through the above algorithm, this invention not only realizes two-dimensional imaging localization of defects, but also realizes quantitative inversion of defect burial depth, thereby improving the level of automation and intelligence in detection.

[0066] Example 5: Comparison with other detection technologies To comprehensively evaluate the performance of this invention, comparative experiments were conducted with existing detection technologies: In summary, this invention has comprehensive advantages in terms of surface adaptability, testing efficiency, safety, and cost, and is particularly suitable for rapid on-site testing of complex curved surface components such as aero-engine blades.

[0067] This invention proposes a method and device for detecting delamination defects in ceramic matrix composites based on a flexible microwave resonator 100. Through flexible conformal bonding, a high-Q open-ended resonator ring, two-dimensional scanning imaging, and frequency offset analysis, it achieves highly sensitive localization, imaging, and quantitative depth inversion of internal delamination defects in ceramic matrix composites. Simulation and experimental results show that this method has excellent surface adaptability, high detection accuracy, and good repeatability. It can detect minute defects as small as 0.5mm × 0.5mm, with a depth inversion error of less than ±0.1mm, and the detection efficiency is more than 5 times higher than traditional methods. This invention solves the technical challenges of rigid sensors being unable to adapt to curved surface detection, low sensitivity to minute defects, and difficulty in quantitative characterization, providing a reliable non-destructive testing method for the widespread application of ceramic matrix composites in aerospace, nuclear energy, and other fields.

[0068] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. 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. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A flexible microwave resonator for detecting delamination defects in ceramic matrix composite materials, characterized in that, include: A flexible microwave resonator includes a flexible dielectric substrate, an open resonant ring fixed on the top of the flexible dielectric substrate, and a coplanar waveguide feeding structure fixed on the bottom of the flexible dielectric substrate. A vector network analyzer is used to transmit microwave signals to the flexible microwave resonator and acquire the S11 curve. A coaxial transmission line connects the flexible microwave resonator to the vector network analyzer; A two-dimensional scanning platform is used to control the flexible microwave resonator to perform two-dimensional scanning along the surface of the ceramic matrix composite material to be tested. The data processing unit is used to construct a scanned image of the material under test based on the resonant frequency of the S11 curve at each scanning point, so as to realize the detection of layered defects.

2. The flexible microwave resonator for detecting delamination defects in ceramic matrix composite materials according to claim 1, characterized in that, The coplanar waveguide feeding structure includes a signal transmission line and grounded copper foil located on both sides of the signal transmission line. A gap is provided between the signal transmission line and the grounded copper foil on both sides, and an SMA connector for connecting a vector network analyzer is provided at one end of the signal transmission line.

3. The flexible microwave resonator for detecting delamination defects in ceramic matrix composite materials according to claim 2, characterized in that, The signal transmission line faces and crosses the central region of the open-circuit resonator, and excites the resonant response of the open-circuit resonator through near-field electromagnetic coupling.

4. The flexible microwave resonator for detecting delamination defects in ceramic matrix composite materials according to claim 1, characterized in that, The flexible dielectric substrate is made of polyimide thin film circuit board.

5. The flexible microwave resonator for detecting delamination defects in ceramic matrix composite materials according to claim 1, characterized in that, The open resonant ring is a copper foil ring with a feeding notch.

6. The flexible microwave resonator for detecting delamination defects in ceramic matrix composite materials according to claim 2, characterized in that, The gap width between the signal transmission line and the grounded copper foil on both sides is 0.05mm~0.2mm; the outer diameter of the open resonant ring is 3mm~10mm, the line width is 0.2mm~0.6mm, and the width of the power supply notch is 0.05mm~0.2mm.

7. A method for detecting delamination defects in ceramic matrix composites based on flexible microwave resonators, applied to the detection device as described in any one of claims 1 to 6, characterized in that, Includes the following steps: Step 1: Apply the flexible microwave resonator with an open resonant ring to the surface of the ceramic matrix composite material to be tested; Step 2: Emitter a microwave signal into the flexible microwave resonator using a vector network analyzer and acquire the S11 curve at the current detection point; Step 3: Extract the resonant frequency from the S11 curve as the microwave signal feature value of the current detection point; Step 4: Control the flexible microwave resonator to perform a two-dimensional scan along the surface of the ceramic matrix composite material to be tested, and repeat steps 2 to 3 to obtain the resonant frequency corresponding to each scanning point; Step 5: Using the coordinates of the scanning points as the position index and the offset of the resonant frequency of the S11 curve as the pixel value, construct a scanned image of the ceramic matrix composite material to be tested, and identify the delamination defect area through image processing algorithms, and invert the defect burial depth based on the frequency offset.

8. The detection method according to claim 7, characterized in that, In step three, if there are layered defects inside the material to be tested, the extracted resonant frequency will shift, and the shift amount will change monotonically with the depth of the defect.

9. The detection method according to claim 7, characterized in that, In step one, the flexible microwave resonator is conformally applied to the surface of the ceramic matrix composite material to be tested, which has a curved structure, by utilizing its flexible properties.

10. The detection method according to claim 7, characterized in that, In step five, the image processing algorithm includes: setting a threshold for the resonant frequency offset, marking regions in the scanned image whose resonant frequency offset exceeds the threshold as layered defect regions; performing connected component analysis on the marked regions to extract the area and location of each defect region; and mapping the average offset of each defect region to the burial depth according to the pre-calibrated frequency offset-burial depth relationship curve.

Images