Epoxy resin bubble microdefect identification method

Abstract

The invention discloses an epoxy resin bubble micro-defect identification method, and belongs to the technical field of epoxy resin micro-defect evaluation, and the method comprises the steps: collecting a transmitted terahertz wave time domain electric signal, carrying out the Fourier transform, and converting the transmitted terahertz wave time domain electric signal to a frequency domain electric signal, obtaining a refractive index and an extinction coefficient, and carrying out the quantitative calculation of a real part and an imaginary part of an optical frequency complex dielectric constant; a fingerprint database of optical frequency dielectric response characteristics of characteristic frequency bands of epoxy resin materials of various preparation processes is established, so that misjudgment caused during defect identification is avoided; calculating a fingerprint range of a comprehensive evaluation index F according to the centralized and standardized optical frequency dielectric response data, a scale factor theta meeting a set cumulative contribution rate and a u-dimensional feature vector; the influence of useless components on analysis is reduced, and the test curve can be quickly analyzed, so that the recognition efficiency is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of epoxy resin micro-defect evaluation, and in particular relates to a method for identifying epoxy resin bubble micro-defects. Background Art

[0002] Epoxy resin insulation, the primary insulation system for dry-type electrical equipment, offers excellent fire and heat resistance, and its insulation strength surpasses that of oil-paper insulation. However, epoxy resin solid insulation can be susceptible to microscopic air bubbles after high-temperature curing. This can cause significant electric field distortion at these locations under strong electric field conditions, posing a significant safety hazard to dry-type electrical equipment. Due to varying curing and preparation processes across manufacturers, vacuum degassing cannot guarantee the complete elimination of air bubbles within epoxy resin materials. Therefore, detecting the presence of microscopic air bubbles within epoxy resin materials under different curing and preparation processes is crucial, as it can significantly improve the insulation reliability of dry-type electrical equipment.

[0003] At present, the traditional detection methods for bubble micro defects in epoxy resin materials mainly include scanning electron microscopy, infrared detection, ultrasonic detection, X-ray detection and electrical performance testing. Among them, the scanning electron microscopy detection method requires destructive sampling of the sample, and the test is extremely random. Since there may be only a few micro defects in the epoxy resin material, it is impossible to obtain global information within the material during sampling; the infrared detection method is more about obtaining new molecular bonds generated after material aging, and cannot obtain information on bubble micro defects for unaged samples; ultrasonic detection requires coupling agent, which limits some detection environments and applications. In addition, in order to be able to detect millimeter-level tiny defects, it is necessary to increase the ultrasonic frequency. While ultrasonic testing has a low sensitivity, its attenuation increases with increasing frequency, leading to sensitivity issues for millimeter-level defects. X-ray testing can fully observe the internal structure of a specimen, but its high-energy particles pose significant safety concerns. When testing electrical properties such as insulation resistance and power-frequency dielectric loss, the micro-defect bubbles within epoxy resin materials are extremely small, making these two electrical parameters less sensitive to localized micro-defects. Breakdown field strength testing often uses a needle-plate electrode structure, resulting in severe field distortion at the needle tip. This can cause insulation breakdown at the needle tip electrode, where no micro-defect bubbles exist, and makes it impossible to guarantee that the needle tip electrode is the location of micro-defects in the specimen. Furthermore, the optical frequency dielectric response of epoxy resin materials exhibits variations under various preparation processes, including different curing agents and curing times, leading to potential misjudgments during defect identification.

[0004] Therefore, a new technical solution is urgently needed in the existing technology to solve this problem. Summary of the Invention

[0005] In order to overcome the deficiencies in the prior art, the present invention provides a method for identifying epoxy resin bubble micro-defects, which is used to solve the problem that bubble micro-defects in epoxy resin cannot be accurately identified and located.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is: a method for identifying epoxy resin bubble micro-defects, comprising the following steps:

[0007] S1. Using different curing methods and different curing agents, prepare epoxy resin samples with various preparation processes; the epoxy resin samples include multiple groups of defect-free epoxy resin samples and multiple groups of epoxy resin samples with bubble micro-defects;

[0008] S2. Using a transmission terahertz time-domain spectroscopy system to collect the transmitted terahertz wave time-domain electrical signals from multiple groups of prepared epoxy resin samples, and converting them into frequency-domain electrical signals after Fourier transformation to obtain spectrum curves of refractive index and extinction coefficient;

[0009] S3. Quantitatively calculate the real part and imaginary part of the optical frequency complex dielectric constant based on the refractive index and extinction coefficient obtained in S2; construct a real part curve of the optical frequency dielectric response and an imaginary part curve of the optical frequency dielectric response;

[0010] S4. Compare the optical frequency dielectric response curves of epoxy resin samples with bubble micro-defects and epoxy resin samples without defects under the same curing method and the same curing agent, and obtain the characteristic frequency band caused by the bubble micro-defects in the sample in the dispersion region of the curve:

[0011] S5. Changing the curing method and the type of curing agent to establish a fingerprint library of the optical frequency dielectric response characteristics of the epoxy resin material in the characteristic frequency band; the fingerprint library includes a set of real part curves and a set of imaginary part curves of the optical frequency dielectric response characteristics in the characteristic frequency band of the epoxy resin material;

[0012] S6. Performing a centering and standardized dimensionality reduction analysis on the characteristic frequency band optical frequency dielectric response data of the epoxy resin material in S5; calculating the contribution rate, scale factor θ, and u-dimensional eigenvector using a covariance matrix based on the centered and standardized optical frequency dielectric response data;

[0013] S7. Calculate the fingerprint range of the comprehensive evaluation index F based on the centralized and standardized optical frequency dielectric response data, the scale factor θ that satisfies the set cumulative contribution rate, and the u-dimensional eigenvector;

[0014] S8. Apply a transmission terahertz time-domain spectroscopy system to scan the actual epoxy resin material point by point to obtain optical frequency dielectric response data. Execute S6 and S7 to obtain the comprehensive evaluation index F of each point, and determine whether the comprehensive evaluation index F is within the fingerprint range to achieve accurate identification of bubble micro-defects in the epoxy resin material.

[0015] Through the above design scheme, the present invention can bring the following beneficial effects:

[0016] 1. Collect the transmitted terahertz wave time-domain electrical signal and convert it into a frequency-domain electrical signal after Fourier transformation. The refractive index and extinction coefficient are obtained for quantitative calculation of the real and imaginary parts of the optical frequency complex dielectric constant. A fingerprint library of the optical frequency dielectric response characteristics of epoxy resin materials with various preparation processes in the characteristic frequency bands is established to avoid misjudgment during defect identification.

[0017] 2. The fingerprint range of the comprehensive evaluation index F is calculated based on the centralized and standardized optical frequency dielectric response data, the scale factor θ that meets the set cumulative contribution rate, and the u-dimensional eigenvector; the influence of useless components on the analysis is reduced, and the test curve can be quickly analyzed, thereby improving the recognition efficiency.

[0018] 3. Apply a transmission terahertz time-domain spectroscopy system to scan the actual epoxy resin material point by point to obtain optical frequency dielectric response data. Execute S6 and S7 to obtain the comprehensive evaluation index F of each point, and determine whether the comprehensive evaluation index F is within the fingerprint range to achieve accurate identification of bubble micro-defects in the epoxy resin material. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a flow chart of a method for identifying epoxy resin bubble micro-defects.

[0020] Figure 2 This is a time domain electrical signal curve diagram of epoxy resin materials with and without bubble micro-defects in an embodiment of a method for identifying epoxy resin bubble micro-defects of the present invention.

[0021] Figure 3 This is a real part curve of the optical frequency dielectric response of epoxy resin materials with and without bubble micro-defects in an embodiment of a method for identifying epoxy resin bubble micro-defects of the present invention.

[0022] Figure 4 This is a graph showing the imaginary part of the optical frequency dielectric response of epoxy resin materials with and without bubble micro-defects in an embodiment of a method for identifying epoxy resin bubble micro-defects of the present invention. DETAILED DESCRIPTION

[0023] The following is a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings.

[0024] To more clearly illustrate the present invention, the present invention is further described below with reference to preferred embodiments. Those skilled in the art will appreciate that the detailed description below is illustrative and non-restrictive, and that the user may make various changes to the following parameters without departing from the mechanism and scope of the invention as set forth in the claims. To avoid obscuring the essence of the present invention, well-known methods and processes are not described in detail.

[0025] For defects inside epoxy resin materials, the present invention mainly targets internal bubble defects. The dielectric properties of the air in the bubbles are significantly different from those of the epoxy resin material, so the complex dielectric constant is used as a characteristic parameter. When terahertz waves pass through epoxy resin materials, they will produce scattering, absorption and other effects, causing changes in the macroscopic time domain spectrum. Epoxy resin materials prepared by various processes have different scattering and absorption characteristics, making it impossible to use a single curve discrimination method to distinguish materials prepared by different processes. Therefore, it is necessary to calculate the complex dielectric constant using the extinction coefficient and refractive index, and establish a fingerprint library of the optical frequency dielectric response characteristics of the characteristic frequency bands of epoxy resin materials prepared by various processes to avoid misjudgment during defect identification.

[0026] By the attached Figures 1 to 4 As shown: A method for identifying epoxy resin bubble micro-defects, comprising the following steps:

[0027] S1. Using different curing methods and different curing agents, prepare epoxy resin samples with various preparation processes; the epoxy resin samples include multiple groups of defect-free epoxy resin samples and multiple groups of epoxy resin samples with bubble micro-defects;

[0028] S2. Using a transmission terahertz time-domain spectroscopy system to collect the transmitted terahertz wave time-domain electrical signals from multiple groups of prepared epoxy resin samples, and converting them into frequency-domain electrical signals after Fourier transformation to obtain spectrum curves of refractive index and extinction coefficient;

[0029] S3. Quantitatively calculate the real part and imaginary part of the optical frequency complex dielectric constant based on the refractive index and extinction coefficient obtained in S2; construct a real part curve of the optical frequency dielectric response and an imaginary part curve of the optical frequency dielectric response;

[0030] S4. Compare the optical frequency dielectric response curves of epoxy resin samples with bubble micro-defects and epoxy resin samples without defects under the same curing method and the same curing agent, and obtain the characteristic frequency band caused by the bubble micro-defects in the sample in the dispersion region of the curve:

[0031] S5. Changing the curing method and the type of curing agent to establish a fingerprint library of the optical frequency dielectric response characteristics of the epoxy resin material in the characteristic frequency band; the fingerprint library includes a set of real part curves and a set of imaginary part curves of the optical frequency dielectric response characteristics in the characteristic frequency band of the epoxy resin material;

[0032] S6. Performing a centering and standardized dimensionality reduction analysis on the characteristic frequency band optical frequency dielectric response data of the epoxy resin material in S5; calculating the contribution rate, scale factor θ, and u-dimensional eigenvector using a covariance matrix based on the centered and standardized optical frequency dielectric response data;

[0033] S7. Calculate the fingerprint range of the comprehensive evaluation index F based on the centralized and standardized optical frequency dielectric response data, the scale factor θ that meets the set cumulative contribution rate, and the u-dimensional eigenvector.

[0034] S8. Perform point-by-point scanning on the actual epoxy resin material using the transmission terahertz time-domain spectroscopy system to obtain the optical frequency dielectric response data. Execute S6 and S7 to obtain the comprehensive evaluation index F at each point, and determine whether the comprehensive evaluation index F is within the fingerprint range to achieve accurate identification of the micro-bubble defects in the epoxy resin material.

[0035] Furthermore, the formulas for calculating the real part and the imaginary part of the optical frequency complex dielectric constant in S3 are as follows:

[0036] ε'(ω) = n 2 [[ID=...]] (ω) - k 2 (ω)

[0037] ε”(ω) = 2n(ω)k(ω)

[0038] In the formula, n is the refractive index; k is the extinction coefficient; ω is the angular frequency.

[0039] The calculation steps of the refractive index and the extinction coefficient are as follows:

[0040] The transfer function H of the actual test sample and the reference sample:

[0041]

[0042] In the formula, E s is the optical frequency time-domain electrical signal passing through the sample; E r is the optical frequency time-domain electrical signal without placing any test items; ω is the angular frequency, unit: radian / second; d is the thickness of a certain layer of the test sample, unit: millimeter; c is the speed of light in vacuum, 3×10 8 m / s; ρ(ω), are the correlation coefficients of the complex refractive index; n~ is the complex refractive index of the material.

[0043]

[0044] In the formula, n is the refractive index; k is the extinction coefficient.

[0045] If the absorption characteristics of the epoxy resin material at the optical frequency of 10 11 ~3×10 12 Hz are very weak, that is, when k < n, the modulus and phase angle of the transfer function H in formula (1):

[0046]

[0047] The parameter meanings in the formula are the same as those in formulas (1) to (3).

[0048] Therefore, the refractive index and extinction coefficient of the epoxy resin material being tested can be obtained:

[0049]

[0050]

[0051] Furthermore, the characteristic frequency band described in S4 is a frequency band with a starting point at a frequency point when the relative error of the optical frequency dielectric response characteristic curve of the epoxy resin sample with bubble micro-defects and the epoxy resin sample without defects is greater than 10%.

[0052] The optical frequency dielectric response data of the characteristic frequency band of the S5 epoxy resin material is further centered and standardized, which can be calculated by the following formula:

[0053]

[0054] Where x(i, j) is the jth eigenvalue of the i-th sample; m is the number of samples, unit: piece; u is the number of characteristics, unit: piece.

[0055] Based on the centered and standardized optical frequency dielectric response data, the steps for calculating the contribution rate, scale factor θ, and u-dimensional eigenvector using the covariance matrix are as follows:

[0056] Calculate the covariance matrix C and find the eigenvalues ​​and eigenvectors of the covariance matrix.

[0057]

[0058] Where x(i) * are all u eigenvalues ​​of the ith sample after central standardization.

[0059]

[0060] Where x(i) * mean is x(i) * The scale factor is obtained by the following formula:

[0061] Cy * (p) = θy * (p)

[0062] y * (P) is the u-dimensional feature vector,

[0063] Calculate the contribution rate, which is the ratio of the eigenvalue of a principal component to the total eigenvalue.

[0064] The contribution rate of the kth eigenvalue is calculated as:

[0065]

[0066] Where λ(k) is the contribution rate of the kth scale factor, and θ(k) is the kth scale factor.

[0067] Furthermore, the cumulative contribution rate range is set to be greater than 95% in S7.

[0068] The formula for calculating the comprehensive evaluation index F in S7 is as follows:

[0069]

[0070] Where x c * The array obtained by centering and standardizing the epoxy resin characteristic data; y * (o) is the u-dimensional eigenvector, o=1…t, t is the total number of scale factors with cumulative contribution rate exceeding 95%, θ(o) is the scale factor; T is the symbol for converting the matrix.

[0071] In the specific implementation, the matrix, curing agent, and catalyst were mixed in a mass ratio of 100:85:1, and five typical curing methods and four typical curing agents were used to prepare epoxy resin samples with various preparation processes. The sample size was 100 mm × 100 mm × 1 mm. The five curing methods were: ① curing at 80 degrees Celsius for 2 hours + curing at 100 degrees Celsius for 2 hours + curing at 120 degrees Celsius for 2 hours, ② curing at 90 degrees Celsius for 3 hours + curing at 110 degrees Celsius for 3 hours, ③ curing at 80 degrees Celsius for 2 hours + curing at 120 degrees Celsius for 4 hours, ④ curing at 80 degrees Celsius for 3 hours + curing at 120 degrees Celsius for 3 hours, and ⑤ curing at 100 degrees Celsius for 4 hours + curing at 120 degrees Celsius for 2 hours. The four different curing agents were: ① methyltetrahydrophthalic anhydride curing agent, ② methylhexahydrophthalic anhydride curing agent, ③ methyltetrahydrophthalic anhydride curing agent, and ④ dicyandiamide curing agent. Twenty epoxy resin samples were obtained by orthogonal experiment. The epoxy resin samples were subjected to 10 11 ~3×10 12 Test of obtaining time domain electrical signals in the Hz range. 10 11 ~3×10 12 The spectrum in the Hertz range is generated by a femtosecond laser, which excites high-frequency electromagnetic waves through a photoconductive antenna. The complete time-domain spectrum change waveform can be obtained through optical rectification sampling and controlling the time delay line.

[0072] The dielectric response data of the characteristic frequency band in the real part and the dielectric response data of the characteristic frequency band in the imaginary part are used as the identification data of epoxy resin micro-defects. Based on the identification characteristic data of 20 epoxy resin samples, they are centered and standardized to obtain the main scale factors and their eigenvectors with a characterization contribution rate of more than 95%, as shown in Tables 1 and 2.

[0073] Table 1 Eigenvectors corresponding to scale factors

[0074]

[0075] Table 2 Scaling factors and contribution

[0076]

[0077] According to Formula 12, the fingerprint range of the comprehensive evaluation index F of micro-defects of 20 epoxy resin materials with different curing processes and curing agents is 13.85±5%.

[0078] To verify the effectiveness of this method, methyltetrahydrophthalic anhydride and methylhexahydrophthalic anhydride were used as curing agents in the present invention. The curing process was 100 degrees Celsius for 8 hours. Two groups of epoxy resin materials with and without defects were prepared. Each group contained 5 epoxy resin samples with and without defects, for a total of 20. A transmission terahertz time-domain spectroscopy system was used to scan the samples point by point to obtain optical frequency dielectric response data. The scanning step length was 0.5 mm.

[0079] The defective epoxy resin material prepared using methyltetrahydrophthalic anhydride as a curing agent and a curing process of 100 degrees Celsius for 8 hours has the following defect location (X, Y):

[0080] (45.0mm, 71.2mm), (31.3mm, 53.3mm), (15.3mm, 17.4mm),

[0081] (3.3mm, 29.4mm), (78.5mm, 69.1mm).

[0082] Execute S6 and S7 to obtain the comprehensive evaluation index F of each point. The F values ​​of the defects are:

[0083] 13.16, 13.56, 14.23, 14.44, 13.56.

[0084] The defective epoxy resin material prepared using methyl hexahydrophthalic anhydride as a curing agent and a curing process of 100 degrees Celsius for 8 hours has the following defect location (X, Y):

[0085] (13.8mm, 40.2mm), (6.2mm, 72.6mm), (36.8mm, 58.0mm),

[0086] (78.4mm, 37.4mm), (35.1mm, 77.5mm).

[0087] Execute S6 and S7 to obtain the comprehensive evaluation index F of each point. The F values ​​of the defects are:

[0088] 13.19, 14.36, 13.99, 13.83, 13.78.

[0089] The comprehensive evaluation index F of the defects is within the fingerprint range, and the evaluation accuracy reaches 100%.

[0090] Obviously, the embodiments described above are only some of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0091] The above mainly introduces the solution of the embodiment of the present application from the perspective of the execution process of the method side. It is understandable that in order to implement the above functions, the network device includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiment provided herein,

[0092] This application can be implemented in hardware or a combination of hardware and computer software. Whether a function is implemented in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0093] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory and includes a number of instructions for enabling a computer device to perform all or part of the steps of the various embodiments of the present application, such as a personal computer, server, or TRP. The aforementioned memory includes various media that can store program codes, such as a USB flash drive, a read-only memory ROM, a random access memory RAM, a mobile hard disk, a magnetic disk, or an optical disk.

Claims

1. A method for identifying epoxy resin bubble micro-defects, characterized in that: The following steps are involved: S1. Using different curing methods and different curing agents, prepare epoxy resin samples with various preparation processes; the epoxy resin samples include multiple groups of defect-free epoxy resin samples and multiple groups of epoxy resin samples with bubble micro-defects; S2. Using a transmission terahertz time-domain spectroscopy system to collect the transmitted terahertz wave time-domain electrical signals from multiple groups of prepared epoxy resin samples, and converting them into frequency-domain electrical signals after Fourier transformation to obtain spectrum curves of refractive index and extinction coefficient; S3. Quantitatively calculate the real part and imaginary part of the optical frequency complex dielectric constant based on the refractive index and extinction coefficient obtained in S2; construct a real part curve of the optical frequency dielectric response and an imaginary part curve of the optical frequency dielectric response; S4. Compare the optical frequency dielectric response curves of epoxy resin samples with bubble micro-defects and epoxy resin samples without defects under the same curing method and the same curing agent, and obtain the characteristic frequency band caused by the bubble micro-defects in the sample in the dispersion region of the curve: S5. Changing the curing method and the type of curing agent to establish a fingerprint library of the optical frequency dielectric response characteristics of the characteristic frequency band of the epoxy resin material; the fingerprint library includes a set of real part curves and a set of imaginary part curves of the optical frequency dielectric response characteristics within the characteristic frequency band of the epoxy resin material; S6. Performing a centering and standardized dimensionality reduction analysis on the optical frequency dielectric response data of the characteristic frequency band of the epoxy resin material in S5; calculating the contribution rate, the scale factor θ, and the u-dimensional eigenvector using a covariance matrix based on the centered and standardized optical frequency dielectric response data; S7. Calculate the fingerprint range of the comprehensive evaluation index F based on the centralized and standardized optical frequency dielectric response data, the scale factor θ that meets the set cumulative contribution rate range, and the u-dimensional eigenvector; S8. Apply a transmission terahertz time-domain spectroscopy system to scan the actual epoxy resin material point by point to obtain optical frequency dielectric response data. Execute S6 and S7 to obtain the comprehensive evaluation index F of each point, and determine whether the comprehensive evaluation index F is within the fingerprint range to achieve accurate identification of bubble micro-defects in the epoxy resin material.

2. The method for identifying epoxy resin bubble micro-defects according to claim 1, wherein: The formula for calculating the real and imaginary parts of the complex dielectric constant at optical frequency described in S3 is as follows: e'(ω)=n 2 (ω)-k 2 (oh) ε”(ω)=2n(ω)k(ω) Where n is the refractive index; k is the extinction coefficient; ω is the angular frequency.

3. The method for identifying epoxy resin bubble micro-defects according to claim 1, wherein: The characteristic frequency band S4 is a frequency band with a starting point at a frequency point when the relative error of the optical frequency dielectric response characteristic curves of the epoxy resin sample with bubble micro-defects and the epoxy resin sample without defects is greater than 10%.

4. The method for identifying epoxy resin bubble micro-defects according to claim 1, wherein: The cumulative contribution rate is set to be greater than 95% in S7.

5. The method for identifying epoxy resin bubble micro-defects according to claim 1, wherein: The formula for calculating the comprehensive evaluation index F in S7 is as follows: Where x c * The array obtained by centering and standardizing the epoxy resin characteristic data; y * (o) is the u-dimensional eigenvector, o=1…t, t is the total number of scale factors with cumulative contribution rate exceeding 95%, θ(o) is the scale factor; T is the symbol for converting the matrix.

Images