Defect detection probability calculation method

Abstract

The invention provides a defect detection probability calculation method which comprises the following steps: S1, selecting a sample piece to be subjected to pore defects, and carrying out conventional CT detection and fine CT detection to obtain a defect conventional detection result and a defect fine detection result, the detection results comprising the number and size information of the pore defects; s2, interval division is carried out on the pore defects according to different pore sizes, the pore defect detection probability of different defect pore size intervals is obtained through the step S1, and the defect detection probability is equal to the ratio of a defect conventional detection result to a defect fine detection result; and S3, taking the intermediate value of the pore size interval as an independent variable, taking the pore defect detection probability as a dependent variable, and obtaining a fitting probability distribution of the pore defect detection probability and the pore size interval. According to the method, defect detection probability information given for pore defects of specific sizes can be obtained.

Description

Technical Field

[0001] The present invention relates to the field of CT detection, and in particular to the field of blade detection. Background Art

[0002] CT inspection (Computed Tomography inspection) refers to the use of computer tomography technology to clearly, accurately and intuitively display the internal structure, composition, material and defect status of the inspected object in the form of two-dimensional tomographic images or three-dimensional stereo images without damaging the inspected object.

[0003] Compared with traditional metal materials, composite materials have the advantages of low density, high specific strength, impact resistance, corrosion resistance, etc., and have broad application prospects in aerospace and other fields. Carbon fiber reinforced resin-based composite materials are widely used in aircraft and engine cold end components due to their obvious weight reduction advantages. Three-dimensional woven composite materials have strong designability and good interlayer performance, and can be applied to fan blades with variable curvature, large thickness, and complex structural shapes. As a key rotating part, the internal quality inspection and evaluation of composite fan blades are particularly important. The most common defect of three-dimensional woven composite blades is pore defects. For pore defects in three-dimensional woven composite materials, the commonly used non-destructive testing methods are ultrasonic testing and radiographic testing. However, ultrasonic testing and radiographic testing can only detect single pore defects of a certain size. It is difficult to quantitatively detect smaller and densely distributed pores, and the severity of the pores can only be qualitatively given based on the attenuation of ultrasound and radiation. CT detection technology can clearly and intuitively display the internal structure, composition, defects and other information of the inspected object in the form of two-dimensional tomographic images or three-dimensional stereo images under non-destructive conditions. Compared with ultrasonic testing and radiographic testing, it can count and measure the internal pore distribution and size of three-dimensional woven composite blades.

[0004] However, existing CT detection methods and image processing software can only perform simple quantitative statistics and size measurements of detected pore defects and display pore distribution. They cannot quantitatively evaluate the detection capability of pore defects, nor can they provide the detection probability (probability of detection) for pores of a specific size. The detection probability refers to the probability of a certain type of defect being detected in non-destructive testing, and thus it is difficult to meet the needs of engine strength design, calculation of the required pore defect distribution and detection probability curve. Summary of the invention

[0005] An object of the present invention is to provide a method for calculating defect detection probability.

[0006] The defect detection probability calculation method for achieving the above purpose includes the following steps: S1. Selecting samples with pore defects, performing conventional CT detection and fine CT detection, and obtaining conventional defect detection results and fine defect detection results, wherein the detection results include the number and size information of the pore defects; S2. Dividing the pore defects into intervals with different pore sizes, and obtaining the pore defect detection probability of different defect pore size intervals through step S1, wherein the defect detection probability is equal to the ratio of the conventional defect detection result to the fine defect detection result; S3. Taking the middle value of the pore size interval as the independent variable and the pore defect detection probability as the dependent variable, obtaining the fitting probability distribution of the pore defect detection probability and the pore size interval.

[0007] In one or more embodiments, in step S1, when performing fine CT detection, the sample is first divided into sections, and then a fine CT scan is performed on each section.

[0008] In one or more embodiments, the CT detection images of each partition are spliced ​​to obtain a complete and high-precision detection result of the sample.

[0009] In one or more embodiments, the parameters of the conventional CT detection are: sampling time 500 ms, magnification 1.45 times, and number of sampled images 1300.

[0010] In one or more embodiments, the parameters of the fine CT detection are: sampling time 1000 ms, magnification 1.82 times, and number of sampled images 1800.

[0011] In one or more embodiments, in step S3, a distribution image of the median value of the pore size interval and the pore defect detection probability is first obtained, and then the distribution image is fitted to obtain a pseudo-fitted probability distribution.

[0012] In one or more embodiments, the sample is a composite blade.

[0013] In one or more embodiments, in step S1, the detection result also includes location information of the pore defect.

[0014] The above method analyzes the CT detection porosity defect data of typical samples, evaluates the detection probability of CT detection porosity defects under different CT detection processes and parameters, and provides a detection probability curve input for composite blade strength evaluation, thereby meeting the needs of engine strength design and calculation of the required porosity defect distribution and detection probability curve. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The above and other features, properties and advantages of the present invention will become more apparent through the following description in conjunction with the accompanying drawings and embodiments, in which:

[0016] Figure 1 It is a flow chart of the defect detection probability calculation method;

[0017] Figure 2A-2B It is a schematic diagram of the partitioning of the blade sample;

[0018] Figure 3 This is a conventional CT inspection image of a composite blade;

[0019] Figure 4 This is a detailed CT inspection image of a composite blade;

[0020] Figure 5 It is the detection probability curve of pore defects;

[0021] Figure 6 It is a flowchart of a specific embodiment of the calculation method. DETAILED DESCRIPTION

[0022] The present invention is further described below in conjunction with specific embodiments and drawings. More details are elaborated in the following description to facilitate a full understanding of the present invention. However, the present invention can obviously be implemented in a variety of other ways different from the description herein. Those skilled in the art can make similar generalizations and deductions based on actual application situations without violating the connotation of the present invention. Therefore, the protection scope of the present invention should not be limited by the content of this specific embodiment.

[0023] It should be noted that these and other subsequent drawings are only examples and are not drawn to scale, and should not be used to limit the actual scope of protection required by the present invention.

[0024] Taking the internal pore defects of three-dimensional woven composite blades as an example, due to the large size of the blades and their complex surface, the defect resolution and detection capabilities are different under different CT detection processes and detection parameters, that is, the detection probability of pore defects is different. How to evaluate the detection probability (probability of detection) of internal pore defects in composite parts under different CT detection processes and detection parameters is a more difficult problem to deal with.

[0025] Existing CT detection methods and image processing software can only perform simple quantitative statistics and size measurements on the detected pore defects and display the pore distribution. They cannot quantitatively evaluate the detection capability of pore defects, nor can they provide a detection probability curve for pores of a specific size. It is difficult to meet the needs of engine strength design and calculation of the pore defect distribution and detection probability curve required.

[0026] Reference Figure 1As shown, the present application discloses a defect detection probability calculation method, comprising the following steps: S1. Selecting a sample to be detected with pore defects, performing conventional CT detection and fine CT detection, obtaining conventional defect detection results and fine defect detection results, the detection results including the number and size information of the pore defects; S2. Dividing the pore defects into intervals with different pore sizes, obtaining the pore defect detection probability of different pore size intervals through step S1, the defect detection probability is equal to the ratio of the conventional defect detection result to the fine defect detection result; S3. Taking the middle value of the pore size interval as the independent variable, and taking the pore defect detection probability as the dependent variable, obtaining the fitting probability distribution of the pore defect detection probability and the pore size interval.

[0027] Samples include but are not limited to composite blades.

[0028] Specifically, in step S1, conventional CT detection refers to performing CT detection on the above-selected sample with typical pore defects under normal production process conditions using a set of fixed parameters to obtain a conventional CT detection image of the blade, such as Figure 3 As shown. Fine CT detection means that the detected image is finer than that of conventional CT detection, and more accurate defect information can be obtained. For example, in some embodiments, conventional process CT detection uses a 450KV|tomeCT detection device with a maximum voltage of 300KV, and fine CT detection uses a 300KV CT detection device with a maximum voltage of 450KV.

[0029] The specific parameters of routine testing are shown in Table 1.

[0030] Table 1 Parameters of conventional CT detection

[0031] Voltage Current Sampling time Magnification Number of sampled images 430kv 1600μA 500ms 1.45 times 1300

[0032] The specific parameters of fine detection are shown in Table 2.

[0033] Table 2 Parameters of fine CT detection

[0034] Voltage Current Sampling time Magnification Number of sampled images 300kv 700μA 1000ms 1.82 times 1800

[0035] Under the fine CT detection equipment and detection parameters with higher resolution and higher detection accuracy, local high-precision CT scanning is performed on different partitions of the above composite blade sample to obtain CT scanning images of each partition. In order to ensure higher resolution and more comprehensive defect detection, the blade can be dissected into several small-sized samples when necessary, such as Figure 2A and 2B As shown in the figure, these small-sized samples are tested by micro-nano CT respectively to obtain the CT test results of pore defects at higher resolution. Finally, the high-precision CT test images of the above-mentioned different partitions or anatomical parts are spliced ​​to obtain the high-precision CT test complete image of the composite blade sample as a whole, as shown in the figure. Figure 4 The high-resolution CT defect detection results can be considered to represent the real defect distribution inside the composite blade sample.

[0036] The defect detection results of step S1 are analyzed to form the statistical results of the pore defects of the samples, as shown in Table III.

[0037] Table 3 Distribution of defect detection results

[0038]

[0039]

[0040] In Table III, the last row N_norm / N_H is the ratio of the defect routine detection result to the defect fine detection result, that is, the defect detection probability.

[0041] Then, the interval Δ is divided according to different pore defect sizes. i =[a i ,a i+1 ], a i is the initial defect size of the interval, a i+1 is the defect size at the end of the interval, comparing different pore defect size intervals△ i The statistical results of pore defects in conventional CT detection process and high-resolution CT defect detection results are used to obtain the percentage N_norm / N_H of the number of pore defects detected by conventional process CT detection to the total number of real pore defects N_H, and the detection probability of pore defects in this size range is obtained.

[0042] The median value of each size interval (a i +a i+1 ) / 2 as the independent variable x, and the detection probability Pi of each size interval as the y value, and draw the detection probability curve of pore defects, such as Figure 5 shown.

[0043] The distribution image obtained above is fitted to determine the defect distribution type of the pore defect CT detection.

[0044] Common distribution types include: Probit distribution, whose expression is P = 1-Φ(f(x)), where f(x) represents the size of the middle value (a i +a i+1) / 2 as the independent variable x, and the detection probability Pi of each size interval as a function of the y value; Logit distribution, its expression is P = exp(f(x)) / (1+exp(f(x))); Cloglog distribution, that is, Weibull distribution, its expression is P = 1-exp(-exp(f(x))); Loglog distribution, its expression is P = exp(-exp(-f(x))).

[0045] According to the detection probability distribution type determined above, a suitable fitting formula is selected to fit the detection probability curve of pore defects, and the expression of the pore defect detection probability Pi with respect to the pore size y=Φ(x) is obtained.

[0046] Below through Figure 6 A specific embodiment of the method described in this application is introduced.

[0047] First, a composite fan blade sample with typical pore defects is selected. Then, a set of fixed parameters is used to perform CT detection on the selected composite fan blade sample with typical pore defects to obtain a conventional CT detection image of the blade, and the conventional CT detection image of the blade is analyzed to obtain the number, size, and location information of pore defects within different size ranges inside the blade, forming the statistical results of pore defects in the conventional CT detection process.

[0048] According to the structural characteristics of the above-mentioned composite blades, combined with the requirements of high-precision CT testing equipment for sample size, the samples are divided into CT test zones. Under CT testing equipment and testing parameters with higher resolution and higher detection accuracy, local high-precision CT scanning is performed on different partitions of the above-mentioned composite blade samples to obtain CT scanning images of each partition. In order to ensure higher resolution and more comprehensive defect detection, the blade can be dissected into several small-sized samples when necessary, and micro-nano CT detection is performed on these small-sized samples respectively to obtain fine CT detection results of pore defects at higher resolution.

[0049] In order to ensure higher resolution and more comprehensive defect detection, the blade can be dissected into several small-sized samples when necessary, and these small-sized samples can be subjected to micro-nano CT detection respectively to obtain CT detection results of pore defects at higher resolution.

[0050] Combining the fine CT test results and conventional CT test results obtained in the above two steps, the statistical results of pore defects of the samples are formed as shown in the table. Comparison of different pore defect size intervals △ iThe statistical results of pore defects detected by conventional CT and high-resolution CT are combined to obtain the percentage of pore defects detected by conventional CT to the total number of real pore defects (that is, pore defects detected by fine CT), and the detection probability Pi of pore defects in this size range is obtained.

[0051] The median value of each size interval (a i +a i+1 ) / 2 as x, and the detection probability Pi of each size interval as the y value. The detection probability curve of pore defects is drawn and fitted to obtain the fitting probability distribution of pore defect detection probability and pore size interval.

[0052] Therefore, the above method is aimed at the shortcomings that the existing CT detection methods and image processing software can only perform simple quantitative statistics and size measurement of the detected pore defects and display the pore distribution, but cannot quantitatively evaluate the detection ability of pore defects, nor can it provide the detection probability curve for pores of a specific size. By analyzing the pore defect data of typical composite blade CT detection, the detection probability of pore defects by CT detection under different CT detection processes and parameters is evaluated, and the detection probability curve input is provided for the strength evaluation of composite blades, which meets the needs of engine strength design and calculation of pore defect distribution and detection probability curve.

[0053] The present application uses specific words to describe the embodiments of the present application. For example, "one embodiment", "an embodiment", and / or "some embodiments" refer to a certain feature, structure or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned twice or multiple times in different positions in this specification does not necessarily refer to the same embodiment. In addition, some features, structures or characteristics in one or more embodiments of the present application can be appropriately combined.

[0054] Although the present invention is disclosed as above with preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present invention. Therefore, any modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope defined by the claims of the present invention.

Claims

1. A defect detection probability calculation method, characterized in that: The steps include: S1. Select samples with pore defects to be detected, perform conventional CT detection and fine CT detection, and obtain conventional defect detection results and fine defect detection results, wherein the detection results include the number and size information of pore defects; S2. Divide the pore defects into intervals of different pore sizes, and obtain the pore defect detection probability of different defect pore size intervals through step S1, wherein the defect detection probability is equal to the ratio of the conventional defect detection result to the fine defect detection result; S3. Taking the middle value of the pore size interval as the independent variable and the pore defect detection probability as the dependent variable, the fitting probability distribution of the pore defect detection probability and the pore size interval is obtained.

2. The defect detection probability calculation method according to claim 1, characterized in that: In step S1, when performing fine CT detection, the sample is first divided into zones, and then fine CT scanning is performed on each zone.

3. The defect detection probability calculation method according to claim 2, characterized in that: The CT inspection images of each partition are spliced ​​together to obtain complete and high-precision inspection results for the sample.

4. The defect detection probability calculation method according to claim 1, characterized in that: The parameters of the conventional CT detection are: sampling time 500 ms, magnification 1.45 times, and number of sampled images 1300.

5. The defect detection probability calculation method according to claim 1 or 4, characterized in that: The parameters of the fine CT detection are: sampling time 1000 ms, magnification 1.82 times, and number of sampled images 1800.

6. The defect detection probability calculation method according to claim 1, characterized in that: In step S3, firstly, a distribution image of the median value of the pore size interval and the pore defect detection probability is obtained, and then the distribution image is fitted to obtain the fitted probability distribution.

7. The defect detection probability calculation method according to claim 1, characterized in that: The sample is a composite material blade.

8. The defect detection probability calculation method according to claim 1, characterized in that: In step S1, the detection result also includes location information of the pore defect.

Images