Thermal barrier coating defect analysis method based on terahertz multi-domain imaging weighted fusion

Through terahertz multi-domain imaging weighted fusion technology, the early internal defects of thermal barrier coating are effectively detected, which improves the evaluation accuracy and reduces the safety risks of aircraft engines.

CN120563483APending Publication Date: 2025-08-29AVIC XIAN AIRCRAFT IND GRP CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202510790855.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-06-13
Publication Date
2025-08-29

AI Technical Summary

Technical Problem

The prior art is difficult to effectively detect early internal defects of thermal barrier coatings, and the failure status evaluation accuracy is low, resulting in aircraft engine safety hazards.

Method used

The terahertz multi-domain imaging weighted fusion method is used to intercept the terahertz reflected wave signals on the surface, interior and lower interface of the thermal barrier coating, calculate the time domain and statistical domain index values, and perform two-dimensional imaging map mapping and weighted fusion. Combined with Canny operator and morphological processing, defect characteristics are identified and failure states are evaluated.

Benefits of technology

It improves the reliability and failure evaluation accuracy of early internal defect detection of thermal barrier coatings, and reduces the risk of major safety accidents.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120563483A_ABST
    Figure CN120563483A_ABST
Patent Text Reader

Abstract

The invention provides a thermal barrier coating defect analysis method based on terahertz multi-domain imaging weighted fusion, which comprises the following steps: according to different positions of defects in a thermal barrier coating, respectively intercepting signals of imaging modes of the surface, the interior, the lower interface and the integral structure of the thermal barrier coating from thermal barrier coating signals; calculating a time domain index value and a statistical domain index value in each signal, arranging the index values into a two-dimensional matrix, and drawing a two-dimensional imaging graph; carrying out fusion imaging on the two-dimensional image by adopting a weighted fusion method to obtain a terahertz fusion image; segmenting and identifying defect features in the terahertz fusion image to obtain defect type and size information; according to the thermal barrier coating falling failure judgment criterion, the failure state of the thermal barrier coating is evaluated according to the defect area proportion; according to the method, the terahertz multi-domain imaging graph is adopted to effectively extract the early internal defect characteristics of the thermal barrier coating, the defect characteristics are enhanced through weighted fusion imaging, and finally early internal defect detection and failure evaluation are achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of aviation terahertz detection technology, and in particular relates to a thermal barrier coating defect analysis method based on terahertz multi-domain imaging weighted fusion. Background Art

[0002] Thermal barrier coatings (TBCs) are composite coating systems consisting of a high-temperature alloy substrate, a metal bonding layer, and a ceramic layer. They are widely used in hot-end components of aircraft engines, gas turbines, and nuclear engineering. Under high-temperature service, a thin thermally grown oxide film forms at the boundary between the metal bonding layer and the ceramic layer within the TBC. As the TBC grows and thickens, microcracks initiate and propagate, merging to form lamellar cracks, causing delamination and peeling of the coating. This ultimately leads to TBC failure, posing a serious threat to engine safety.

[0003] Because thermal barrier coatings are porous and multiphase, and because the ceramic upper and lower interfaces are relatively rough, terahertz signals are severely attenuated during transmission, making internal defect signatures very weak. Current terahertz imaging methods primarily rely on maximum peak amplitude imaging, which typically extracts structural features near the ceramic upper surface, making it difficult to detect early-stage internal defects in thermal barrier coatings. Furthermore, single-source amplitude imaging methods are susceptible to interference from factors such as environmental noise, system perturbations, and fluctuations in terahertz source power, leading to significant detection errors.

[0004] Therefore, developing a new method for quantitative evaluation of early internal defects of thermal barrier coatings and failure state assessment will have important social and significant economic benefits for achieving quality evaluation and non-destructive testing of thermal barrier coatings, judging the failure state of thermal barrier coatings in advance, and avoiding major safety accidents of turbine blades and aircraft engines. Summary of the Invention

[0005] Purpose of the invention: To provide a thermal barrier coating defect analysis method based on terahertz multi-domain imaging weighted fusion to solve the problems in the prior art of difficult detection of early internal defects of thermal barrier coatings and low accuracy of failure state assessment. The present application provides a thermal barrier coating defect analysis method based on terahertz multi-domain imaging weighted fusion, the method comprising: According to the different locations of defects in the thermal barrier coating, the signals of the thermal barrier coating surface and interior, lower interface and overall structure imaging modes are intercepted from the thermal barrier coating signal; Calculate the time domain index value and the statistical domain index value of each signal, arrange the index values ​​into a two-dimensional matrix, and draw them into a two-dimensional imaging diagram; The two-dimensional imaging images are fused using a weighted fusion method to obtain a terahertz fusion image; defect features in the terahertz fusion image are segmented and identified to obtain defect type and size information.

[0006] Preferably, before intercepting the signals of the imaging modes of the surface and interior, lower interface and overall structure of the thermal barrier coating from the thermal barrier coating signal according to the different locations of the defects in the thermal barrier coating, the method further includes: A terahertz system was used to collect thermal barrier coating signals after thermal cycling experiments.

[0007] Preferably, the method of intercepting the imaging modes of the surface and interior, lower interface and overall structure of the thermal barrier coating from the thermal barrier coating signal according to the different locations of the defects in the thermal barrier coating comprises: According to the different positions of defects in the thermal barrier coating, the first reflected wave signal, the second reflected wave signal and the entire reflected wave signal of the terahertz reflected wave signal are respectively intercepted from the thermal barrier coating signal as signals of the imaging mode of the surface and interior, lower interface and overall structure of the thermal barrier coating.

[0008] Preferably, the calculating of the time domain index value and the statistical domain index value in each signal, arranging the index values ​​into a two-dimensional matrix, and drawing them into a two-dimensional imaging diagram includes: The time domain index values ​​and statistical domain index values ​​of each signal are calculated, totaling 18 types. The index values ​​are arranged into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment, and drawn into a two-dimensional imaging diagram, which constitutes the terahertz imaging diagram of different domain indicators.

[0009] Preferably, the method of using a weighted fusion method to fuse the two-dimensional imaging images to obtain a terahertz fused image includes: The weighted fusion method is used to fuse 18 two-dimensional images to obtain a terahertz fusion image.

[0010] Preferably, the segmenting and identifying defect features in the terahertz fusion image to obtain defect type and size information includes: The Canny operator and morphological processing are used to segment and identify defect features in the terahertz fusion image to obtain defect type and size information.

[0011] Preferably, the method further comprises: According to the failure judgment criteria of thermal barrier coating shedding, the failure state of thermal barrier coating is evaluated by the defect area ratio.

[0012] Preferably, the method of fusing 18 two-dimensional images using a weighted fusion method to obtain a terahertz fused image includes: The image quality of 18 index imaging images was evaluated using the image quality assessment functions NIQE, BRISQUE, and PIQE in MATLAB; Assume that i The picture is Image i ,i =1,2,…,18, for any point on the image p =( x , y ), its pixels can be decomposed into three-channel color values ​​of R, G, and B; The NIQE value of each image is used as a weight parameter and normalized to I nique , perform weighted fusion; in the fusion process, first extract the R, G, B color values ​​of each pixel in the image, and then compare the color values ​​of different images with the corresponding (1- I nique ) are multiplied and superimposed, and then the superimposed R, G, and B color values ​​are combined into a color image to obtain the NIQE weighted fusion image; Following the same process, calculate the weight parameters I brisque and I piqe , and obtain the corresponding fused image; Finally, the NIQE weighted fusion image, the BRISQUE weighted fusion image, and the PIQE weighted fusion image are fused according to weight coefficients of 0.3, 0.4, and 0.3 to obtain the final weighted terahertz fusion image.

[0013] Preferably, NIQE is a non-reference image evaluation index, the smaller its value, the better the perceived quality of the image; BRISQUE is a blind image spatial quality evaluator, the smaller its value, the better the spatial quality evaluation of the image; PIQE is a perception-based image quality evaluator, the lower its score, the better the image quality.

[0014] The beneficial effects of this application are: The present application provides a method for defect detection and failure assessment of thermal barrier coatings based on weighted fusion of terahertz multi-domain imaging. The method uses terahertz multi-domain imaging to effectively extract early internal defect characteristics of thermal barrier coatings, enhances the defect characteristics through weighted fusion imaging, and finally realizes early internal defect detection and failure assessment.

[0015] This method can improve the reliability of early internal defect detection of thermal barrier coatings and the accuracy of failure assessment. It has broad application prospects in detecting internal defects of thermal barrier coatings, analyzing the processing quality of thermal barrier coatings, determining the failure state of thermal barrier coatings, and reducing the risk of major safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a schematic diagram of a terahertz multi-domain imaging weighted fusion detection method for thermal barrier coating defects provided in an embodiment of the present application; Figure 2This is a screenshot of different reflected waves in a terahertz reflected wave signal of a thermal barrier coating provided in an embodiment of the present application; Figure 3 This is a multi-domain index imaging image and a weighted fusion imaging image of a thermal barrier coating in different structural modes after 100 hours of thermal cycling provided in an embodiment of the present application; Figure 4 This is a diagram for identifying and quantitatively evaluating thermal barrier coating defects provided in an embodiment of the present application; Figure 5 This is a thermal barrier coating failure state determination diagram provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] The present application belongs to the field of aviation terahertz detection, and specifically designs a thermal barrier coating defect detection and failure assessment method based on terahertz multi-domain imaging weighted fusion. The method uses a terahertz system to collect thermal barrier coating signals after a thermal cycle experiment, and intercepts terahertz wave signals of different structural modes according to the different positions of defects in the thermal barrier coating; calculates time domain and statistical domain indicators in each signal, and plots the indicator values ​​into terahertz imaging images of different domain indicators; performs weight calculation and weighted fusion on the multi-domain imaging images to obtain terahertz fusion images of different structural modes; uses the Canny operator and morphological processing to segment, identify and quantitatively evaluate the defect features in the image, and uses the delamination defect area ratio to evaluate the failure state of the thermal barrier coating according to the thermal barrier coating shedding failure criterion; can effectively detect early internal defects of the thermal barrier coating and realize failure assessment, thereby improving the reliability of early internal defect detection of the thermal barrier coating and the accuracy of failure assessment.

[0018] This application provides a thermal barrier coating defect analysis method based on terahertz multi-domain imaging weighted fusion, namely a detection and failure assessment method, including: Step 1: Using a terahertz system to collect the thermal barrier coating signal after the thermal cycling experiment, according to the different locations of the defects in the thermal barrier coating, the first reflected wave signal, the second reflected wave signal, and the entire reflected wave signal of the terahertz reflected wave signal are respectively intercepted from the thermal barrier coating signal as signals of the imaging mode of the surface and interior, lower interface, and overall structure of the thermal barrier coating; Step 2: Calculate the time domain index value and statistical domain index value of each signal, a total of 18 types, arrange the index values ​​into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment, and draw a two-dimensional imaging diagram, that is, a terahertz imaging diagram of different domain indicators; Step 3: Use the weighted fusion method to fuse the 18 index images to obtain a terahertz fusion image; use the Canny operator and morphological processing to segment and identify the defect features in the terahertz image to obtain the defect type and size information; according to the thermal barrier coating failure judgment criteria, the failure status of the thermal barrier coating is evaluated by the defect area ratio.

[0019] Wherein, in step 1, the method for intercepting the terahertz signal of the thermal barrier coating in different structural imaging modes is: The entire reflected wave signal was used as the signal for the thermal barrier coating's overall structural imaging mode. The first reflected wave signal was intercepted as the signal for the thermal barrier coating's internal and surface structural imaging mode: based on the time point of the first reflected wave's maximum peak amplitude, a time domain segment of 5 ps was intercepted forward and 3 ps backward, for a total of 8 ps of signal length as the signal for the surface and internal structural imaging mode. The second reflected echo signal was intercepted as the signal for the internal interface structural imaging mode: based on the time point of the second reflected wave's maximum peak amplitude, a time domain segment of 4 ps was intercepted forward and 2 ps backward, for a total of 6 ps of signal length as the signal for the internal interface structural imaging mode.

[0020] Wherein, in step 2, the terahertz multi-domain imaging method of different structural modes of the thermal barrier coating is: The names and calculation formulas of the indicators in the time domain and statistical domain (18 in total) are as follows:

[0021] In the above table, n is the number of signal points, Sample signal f The average amplitude of t is the flight time of the terahertz wave, n is the number of terahertz signal points, N is the total number of points collected by the terahertz signal, which is 4096 in the embodiment of the present application. f ( t )and f ( n ) both represent terahertz time-domain signals, the difference being that the independent variables are the terahertz flight time and the number of terahertz signal points, respectively.

[0022] The index values ​​calculated from signals of different structural modes are arranged into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment, and then plotted into a two-dimensional imaging diagram, which constitutes a terahertz imaging diagram with different indicators in multiple domains (a total of 18 images).

[0023] Wherein, in step 3, the weighted fusion method based on multi-domain imaging images includes: The image quality of 18 index images is evaluated by the image quality assessment functions niqe, brisque, and piqe in MATLAB. NIQE is a non-reference image evaluation index. The smaller its value, the better the perceived quality of the image. BRISQUE is a blind image spatial quality evaluator. The smaller its value, the better the spatial quality evaluation of the image. PIQE is a perception-based image quality evaluator. The lower its score, the better the image quality. Assume that i The picture is Image i ( i =1,2,…,18), for any point on the image p =( x , y ), each pixel can be decomposed into three-channel color values ​​of R, G, and B. The niqe value of each image is used as a weight parameter and normalized to I nique , perform weighted fusion. In the fusion process, first extract the R, G, and B color values ​​of each pixel in the image, and then compare the color values ​​of different images with the corresponding (1- I nique ) are multiplied and superimposed, and then the superimposed R, G, and B color values ​​are combined into a color image to obtain the NIQE weighted fusion image. Following the same process, the weight parameters are calculated. I brisque and I piqe , and obtain the corresponding fused image. Finally, the NIQE weighted fusion image, BRISQUE weighted fusion image, and PIQE weighted fusion image are fused with weight coefficients of 0.3, 0.4, and 0.3 to obtain the final weighted fusion image.

[0024] Among them, in step 3, the thermal barrier coating defect feature segmentation, identification and failure assessment method includes: The Canny operator was used to segment defect features in the terahertz (THz) fusion image. The aspect ratio of the thermal barrier coating (TBC) delamination defect feature was defined as less than or equal to 6, while the aspect ratio of the crack defect feature was defined as greater than 6 to identify the defect type in the THz image. Based on the scanning step size (200 μm) used in the THz scanning imaging experiment, the distance between two adjacent pixels in the THz image was set to 200 μm. The defect size was then calculated based on the number of pixels contained in each defect feature. Failure of the TBC sample was defined as failure when the ceramic delamination area was greater than or equal to 30%. The delamination defect area was set to be equivalent to the ceramic delamination area. The failure state of the TBC was evaluated based on the ceramic delamination area or the delamination defect area.

[0025] In other embodiments of the present application, a method for thermal barrier coating defect detection and failure assessment based on terahertz multi-domain imaging weighted fusion is provided, comprising: Step 1: Use a terahertz system to collect the thermal barrier coating signal after the thermal cycling experiment. According to the different locations of the defects in the thermal barrier coating, the first reflected wave, the second reflected wave, and the entire reflected wave signal of the terahertz reflected wave signal are intercepted as signals for the imaging mode of the thermal barrier coating surface and interior, lower interface, and overall structure; Step 2: Calculate the time domain and statistical domain indicators (a total of 18) for each signal, arrange the indicator values ​​into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment, and draw a two-dimensional imaging diagram, which constitutes the terahertz imaging diagram of different domain indicators; Step 3: A weighted fusion method was used to fuse the 18 index images to produce a terahertz fusion image. The Canny operator and morphological processing were used to segment and identify defect features in the terahertz image, determining defect type and size. Based on the thermal barrier coating failure criteria, the defect area percentage was used to assess the failure state of the thermal barrier coating.

[0026] In at least one embodiment of the present application, in step 1, the method for intercepting terahertz signals in different structural imaging modes of the thermal barrier coating includes: Because the terahertz signal changes caused by defects on and near the ceramic upper surface in the thermal barrier coating structure are primarily reflected in the first reflected wave (i.e., the air-ceramic interface reflection wave), the first reflected wave signal is intercepted as the signal for the thermal barrier coating surface and internal structure imaging mode: based on the time point of the first reflected wave's maximum peak amplitude, a time domain signal of 8 ps is captured forward 5 ps and backward 3 ps, for a total of 8 ps. The terahertz signal changes caused by defects near the ceramic lower interface are primarily reflected in the second reflected wave (i.e., the ceramic-metal bonding layer interface reflection wave). The second reflected wave signal is intercepted as the signal for the internal interface structure imaging mode: based on the time point of the second reflected wave's maximum peak amplitude, a time domain signal of 6 ps is captured forward 4 ps and backward 2 ps, for a total of 6 ps. The entire reflected wave signal is used as the signal for the thermal barrier coating's overall structure imaging mode.

[0027] The index values ​​calculated from signals of different structural modes are arranged into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment, and then plotted into a two-dimensional imaging diagram, which constitutes a terahertz imaging diagram with different indicators in multiple domains (a total of 18 images).

[0028] In at least one embodiment of the present application, in step 3, the weighted fusion method based on the multi-domain imaging image is: The image quality of 18 index images is evaluated by the image quality assessment functions niqe, brisque, and piqe in MATLAB. NIQE is a non-reference image evaluation index. The smaller its value, the better the perceived quality of the image. BRISQUE is a blind image spatial quality evaluator. The smaller its value, the better the spatial quality evaluation of the image. PIQE is a perception-based image quality evaluator. The lower its score, the better the image quality. Assume that i The picture is Image i ( i =1,2,…,18), for any point on the image p =( x , y ), each pixel can be decomposed into three-channel color values ​​of R, G, and B. The niqe value of each image is used as a weight parameter and normalized to I nique , perform weighted fusion. In the fusion process, first extract the R, G, and B color values ​​of each pixel in the image, and then compare the color values ​​of different images with the corresponding (1- I nique ) are multiplied and superimposed, and then the superimposed R, G, and B color values ​​are combined into a color image to obtain the NIQE weighted fusion image. Following the same process, the weight parameters are calculated. I brisque and I piqe , and obtain the corresponding fused image. Finally, the NIQE weighted fusion image, BRISQUE weighted fusion image, and PIQE weighted fusion image are fused with weight coefficients of 0.3, 0.4, and 0.3 to obtain the final weighted fusion image.

[0029] In at least one embodiment of the present application, in step three, the thermal barrier coating defect feature segmentation, identification, and failure assessment method is: The Canny operator was used to segment defect features in the terahertz (THz) fusion image. The aspect ratio of the thermal barrier coating (TBC) delamination defect feature was defined as less than or equal to 6, while the aspect ratio of the crack defect feature was defined as greater than 6 to identify the defect type in the THz image. Based on the scanning step size (200 μm) used in the THz scanning imaging experiment, the distance between two adjacent pixels in the THz image was set to 200 μm. The defect size was then calculated based on the number of pixels contained in each defect feature. Failure of the TBC sample was defined as failure when the ceramic delamination area was greater than or equal to 30%. The delamination defect area was set to be equivalent to the ceramic delamination area. The failure state of the TBC was evaluated based on the ceramic delamination area or the delamination defect area.

[0030] This application provides a method for thermal barrier coating (TBC) defect detection and failure assessment based on weighted fusion of terahertz multi-domain imaging. This method uses terahertz multi-domain imaging to effectively extract early internal defect features of TBCs, enhances these features through weighted fusion imaging, and ultimately enables early internal defect detection and failure assessment. This method improves the reliability of early internal defect detection and the accuracy of failure assessment for TBCs, and has broad application prospects in detecting internal defects in TBCs, analyzing TBC processing quality, determining TBC failure states, and reducing the risk of major safety accidents.

[0031] In other embodiments of the present application, the following Figure 1 -Attached Figure 5 This application is described in further detail.

[0032] This application provides a thermal barrier coating defect detection and failure assessment method based on terahertz total variation fusion imaging, such as Figure 1 As shown, the following steps are included: Step 1: Use a terahertz system to collect the thermal barrier coating signal after the thermal cycling experiment. According to the different locations of the defects in the thermal barrier coating, the first reflected wave, the second reflected wave, and the entire reflected wave signal of the terahertz reflected wave signal are intercepted as signals for the imaging mode of the thermal barrier coating surface and interior, lower interface, and overall structure; Step 2: Calculate the time domain and statistical domain indicators (a total of 18) for each signal, arrange the indicator values ​​into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment, and draw a two-dimensional imaging diagram, which constitutes the terahertz imaging diagram of different domain indicators; Step 3: A weighted fusion method was used to fuse the 18 index images to produce a terahertz fusion image. The Canny operator and morphological processing were used to segment and identify defect features in the terahertz image, determining defect type and size. Based on the thermal barrier coating failure criteria, the defect area percentage was used to assess the failure state of the thermal barrier coating.

[0033] In step 1, the terahertz wave signals of different structural modes of the thermal barrier coating are intercepted. In this embodiment, Figure 2 As shown, the interception method of the terahertz wave signal is: The entire terahertz reflected wave signal was used as the signal for the overall structure imaging mode. The first reflected wave signal was used as the signal for the surface and internal structure imaging mode. Based on the time point of the first reflected wave's maximum peak amplitude, a time domain of 5 ps forward and 3 ps backward was intercepted, for a total of 8 ps of signal length as the signal for the surface and internal structure imaging mode. The second reflected wave signal was used as the signal for the internal interface structure imaging mode: Based on the time point of the second reflected wave's maximum peak amplitude, a time domain of 4 ps forward and 2 ps backward was intercepted, for a total of 6 ps of signal length as the signal for the internal interface structure imaging mode.

[0034] The index values ​​of the thermal barrier coating in the time domain and statistical domain after 100 hours of thermal cycling (a total of 18 types) were calculated, the index values ​​were normalized, and arranged into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment. The Rainbow-Balance color card was selected, and a corresponding relationship was established between the normalized index value range and the color value range of the color card. The index values ​​were then converted into color card values ​​according to the corresponding relationship, thus forming terahertz imaging images of different domain indicators. 18 index imaging images of the overall structure mode (the entire reflected wave), the surface and internal structure mode (the first reflected wave), and the internal interface structure mode (the second reflected wave) were obtained, which are divided into Image 1, Image 2, Image 3,…, Image 18 .

[0035] In step 3, based on the weighted fusion and failure assessment method of multi-domain imaging, in this embodiment, Figure 3 As shown in the figure, the weighted fusion method based on multi-domain imaging is: The image quality of 18 index images is evaluated by the image quality assessment functions niqe, brisque, and piqe in MATLAB. NIQE is a non-reference image evaluation index. The smaller its value, the better the perceived quality of the image. BRISQUE is a blind image spatial quality evaluator. The smaller its value, the better the spatial quality evaluation of the image. PIQE is a perception-based image quality evaluator. The lower its score, the better the image quality. Assume that i The picture is Image i ( i =1,2,…,18), for any point on the image p =( x , y ), each pixel can be decomposed into three-channel color values ​​of R, G, and B. The niqe value of each image is used as a weight parameter and normalized to I nique, perform weighted fusion. In the fusion process, first extract the R, G, and B color values ​​of each pixel in the image, and then compare the color values ​​of different images with the corresponding (1- I nique ) are multiplied and superimposed, and then the superimposed R, G, and B color values ​​are combined into a color image to obtain the NIQE weighted fusion image. Following the same process, the weight parameters are calculated. I brisque and I piqe , and obtain the corresponding fused image. Finally, the NIQE weighted fusion image, BRISQUE weighted fusion image, and PIQE weighted fusion image are fused with weight coefficients of 0.3, 0.4, and 0.3 to obtain the final weighted fusion image.

[0036] Multi-domain imaging of a thermal barrier coating after 100 hours of thermal cycling. After weighted fusion processing, fused images were obtained under different structural imaging modes. A mixture of crack and delamination defects was observed in the overall structural fusion image, crack defects were observed in the surface and internal fusion images, and delamination defects were observed in the internal interface fusion image. This demonstrates that the proposed method can effectively detect early-stage internal defects in thermal barrier coatings, including surface and internal crack defects, as well as delamination defects near internal interfaces.

[0037] In step 3, based on the weighted fusion and failure assessment method of multi-domain imaging, in this embodiment, Figure 4 As shown, the defect feature segmentation, identification and quantitative evaluation method of the thermal barrier coating fusion imaging image is as follows: The Canny operator was used to segment defect features in the fused terahertz (THz) image. Delamination defects in thermal barrier coatings were defined as having an aspect ratio less than or equal to 6, while crack defects were defined as having an aspect ratio greater than 6. The defect types in the THz image were identified based on the morphology of the defect features. Based on the scanning step size (200 μm) used in the THz imaging experiment, the distance between two adjacent pixels in the THz image was set to 200 μm. The defect size was then calculated based on the number of pixels contained in each defect feature.

[0038] Since the internal delamination defect features in the overall structure fusion map are not fully revealed, more detailed crack and delamination defect features are reflected in the surface, internal and internal interface fusion maps respectively. Therefore, the Canny operator and defect morphology definition are used to segment and identify the defect features in the overall structure fusion map of the thermal barrier coating, and the defect features in the surface, internal and internal interface fusion maps are segmented, identified and quantitatively evaluated. It is defined that the aspect ratio of the delamination defect feature of the thermal barrier coating is less than or equal to 6, and the aspect ratio of the crack defect feature is greater than 6. The crack defect is visualized in red and the delamination defect is visualized in yellow. According to the distance between two adjacent pixels, after 80 h to 100 h of thermal cycling, it is calculated that the length of the crack defect expands from 6.83 mm to 8.24 mm, and the area of ​​the delamination defect increases from 21.72 mm to 3. 2 Expanded to 40.47 mm 2 .

[0039] In step 3, based on the weighted fusion and failure assessment method of multi-domain imaging, in this embodiment, Figure 5 As shown, the failure assessment method of thermal barrier coating is: It is defined that shedding failure occurs when the ceramic shedding area of ​​the thermal barrier coating sample is greater than or equal to 30%. At the same time, the area of ​​the delamination defect is set to be equivalent to the ceramic shedding area. The failure state of the thermal barrier coating is evaluated by the area of ​​the ceramic shedding area or the area of ​​the delamination defect.

[0040] After 80 h and 100 h of thermal cycling, the delamination defect area ratio of the thermal barrier coating was calculated to be 14.38% and 26.6%, respectively. At this time, the thermal barrier coating had not yet fallen off and failed, realizing the evaluation of the failure state of the thermal barrier coating.

[0041] The present application provides a thermal barrier coating defect detection and failure assessment method based on terahertz multi-domain imaging weighted fusion. The method utilizes a terahertz system to collect thermal barrier coating signals after a thermal cycling experiment, and intercepts terahertz wave signals of different structural modes according to the different locations of defects in the thermal barrier coating. The method calculates time domain and statistical domain indicators in each signal, and plots the indicator values ​​into terahertz imaging images of different domain indicators. The multi-domain imaging images are weighted and weightedly fused to obtain terahertz fused images of different structural modes. The defect features in the image are segmented, identified, and quantitatively evaluated through the Canny operator and morphological processing. The failure state of the thermal barrier coating is evaluated using the delamination defect area ratio according to the thermal barrier coating shedding failure criterion. The method can effectively detect early internal defects of the thermal barrier coating and implement failure assessment, thereby improving the reliability of early internal defect detection and the accuracy of failure assessment of the thermal barrier coating.

Claims

1. A thermal barrier coating defect analysis method based on terahertz multi-domain imaging weighted fusion, characterized in that: The method comprises: According to the different locations of defects in the thermal barrier coating, the signals of the thermal barrier coating surface and interior, lower interface and overall structure imaging modes are intercepted from the thermal barrier coating signal; Calculate the time domain index value and the statistical domain index value of each signal, arrange the index values ​​into a two-dimensional matrix, and draw them into a two-dimensional imaging diagram; The two-dimensional imaging images are fused using a weighted fusion method to obtain a terahertz fusion image; defect features in the terahertz fusion image are segmented and identified to obtain defect type and size information.

2. The method according to claim 1, wherein Before respectively intercepting the signals of the imaging modes of the surface and interior, lower interface and overall structure of the thermal barrier coating from the thermal barrier coating signals according to the different locations of the defects in the thermal barrier coating, the method further includes: A terahertz system was used to collect thermal barrier coating signals after thermal cycling experiments.

3. The method according to claim 2, wherein The method of intercepting the imaging mode signals of the surface and interior of the thermal barrier coating, the lower interface and the overall structure of the thermal barrier coating from the thermal barrier coating signal according to the different locations of the defects in the thermal barrier coating includes: According to the different positions of defects in the thermal barrier coating, the first reflected wave signal, the second reflected wave signal and the entire reflected wave signal of the terahertz reflected wave signal are respectively intercepted from the thermal barrier coating signal as signals of the imaging mode of the surface and interior, lower interface and overall structure of the thermal barrier coating.

4. The method according to claim 1, wherein The step of calculating the time domain index value and the statistical domain index value in each signal, arranging the index values ​​into a two-dimensional matrix, and drawing the two-dimensional imaging graph includes: The time domain index values ​​and statistical domain index values ​​of each signal are calculated, totaling 18 types. The index values ​​are arranged into a two-dimensional matrix according to the signal acquisition order during the terahertz scanning experiment, and drawn into a two-dimensional imaging diagram, which constitutes the terahertz imaging diagram of different domain indicators.

5. The method according to claim 4, wherein The method of using a weighted fusion method to fuse the two-dimensional imaging images to obtain a terahertz fused image includes: The weighted fusion method is used to fuse 18 two-dimensional images to obtain a terahertz fusion image.

6. The method according to claim 5, wherein The segmenting and identifying defect features in the terahertz fusion image to obtain defect type and size information includes: The Canny operator and morphological processing are used to segment and identify defect features in the terahertz fusion image to obtain defect type and size information.

7. The method according to claim 1, wherein The method further comprises: According to the failure judgment criteria of thermal barrier coating shedding, the failure state of thermal barrier coating is evaluated by the defect area ratio.

8. The method according to claim 5, wherein The method of using a weighted fusion method to fuse 18 two-dimensional images to obtain a terahertz fused image includes: The image quality of 18 index imaging images was evaluated using the image quality assessment functions NIQE, BRISQUE, and PIQE in MATLAB; Assume that i The picture is Image i , i =1,2,…,18, for any point on the image p =( x , y ), its pixels can be decomposed into three-channel color values ​​of R, G, and B; The NIQE value of each image is used as a weight parameter and normalized to I nique , perform weighted fusion; in the fusion process, first extract the R, G, B color values ​​of each pixel in the image, and then compare the color values ​​of different images with the corresponding (1- I nique ) are multiplied and superimposed, and then the superimposed R, G, and B color values ​​are combined into a color image to obtain the NIQE weighted fusion image; Following the same process, calculate the weight parameters I brisque and I piqe , and obtain the corresponding fused image; Finally, the NIQE weighted fusion image, the BRISQUE weighted fusion image, and the PIQE weighted fusion image are fused according to weight coefficients of 0.3, 0.4, and 0.3 to obtain the final weighted terahertz fusion image.

9. The method according to claim 8, wherein NIQE is a no-reference image evaluation index. The smaller its value, the better the perceived quality of the image. BRISQUE is a blind image spatial quality evaluator. The smaller its value, the better the spatial quality evaluation of the image. PIQE is a perception-based image quality evaluator. The lower its score, the better the image quality.

Citation Information

Cited By

  • Nuclear power sealing gasket production quality detection method based on machine vision

    CN121767367A

  • Method, device and storage medium for quantitatively monitoring thermal barrier coating cmas damage spallation

    CN122391651A