A method for detecting surface defects in epoxy-coated aluminum alloy plates.

Abstract

This invention relates to the technical fields of aircraft component processing, testing, and inspection, and provides a method for detecting surface defects in epoxy-coated aluminum alloy plates. The method involves applying an epoxy coating containing zinc yellow pigment onto an anodic oxide film to construct a coated test piece with color difference indication function, which is then exposed to a coastal atmospheric environment for sampling. During the detection process, data such as gloss loss rate, color difference, contact angle, defect area ratio, and corrosion current density are collected. A dynamic correction coefficient is calculated based on the degree of zinc yellow pigment hydrolysis, adaptively adjusting the threshold values ​​of each parameter and constructing a threshold set. Combining scanning electron microscopy image segmentation and Tafel extrapolation analysis, multi-source indicators are integrated to determine whether the coating has failed. This method can track the hydrolysis behavior of zinc yellow pigment in real time, automatically correct the judgment criteria for optical, morphological, and electrochemical indicators, and improve the accuracy and reliability of failure detection of epoxy-coated aluminum alloy plates under coastal service conditions.

Description

Technical Field

[0001] This invention relates to the technical fields of aircraft component processing, testing, and inspection, and in particular to a method for detecting surface defects in epoxy-coated aluminum alloy plates. Background Technology

[0002] Aluminum alloys are widely used in aerospace equipment. To improve corrosion resistance, aluminum alloy surfaces are often anodized first, followed by an epoxy-based protective coating. However, coastal atmospheres are saline, humid, and have strong ultraviolet radiation, making coatings prone to defects such as loss of gloss, discoloration, cracking, and blistering during long-term service. Once a defect penetrates, it can rapidly induce pitting corrosion or even intergranular corrosion in the substrate. Existing defect assessment methods mainly rely on a few indicators (such as gloss loss rate or polarization current) for judgment. These criteria are fixed and cannot reflect the performance fluctuations of the coating caused by environmental changes or self-healing reactions in real time, leading to a high false positive rate and a high risk of leak detection. In addition, traditional color difference monitoring ignores the chemical effects of self-passivating pigments in the coating and cannot distinguish the influence of external fading from internal chemical repair. Although multi-parameter models integrating electron microscopy, image segmentation, and electrochemical techniques have been proposed, they lack a dynamic threshold update mechanism coupled with the coating's self-healing chemical process, making it difficult to adapt to the judgment requirements of different aging stages. Therefore, there is an urgent need for a comprehensive detection method that can track the hydrolysis behavior of zinc yellow pigment in real time, automatically correct multiple thresholds, and integrate optical, morphological, and electrochemical information to accurately assess the failure risk of epoxy-coated aluminum alloy plates under coastal service conditions. Summary of the Invention

[0003] To address the shortcomings of the existing technologies, this invention provides a method for detecting surface defects in epoxy-coated aluminum alloy plates. This method can track the hydrolysis behavior of zinc yellow pigment in real time, automatically correct multiple thresholds, and integrate optical, morphological, and electrochemical information to accurately assess the failure risk of epoxy-coated aluminum alloy plates under coastal service conditions.

[0004] The present invention provides a method for detecting surface defects in epoxy-coated aluminum alloy plates, comprising:

[0005] Anodizing is performed on the aluminum alloy plate substrate to form an oxide film, and an epoxy resin coating containing zinc yellow pigment is applied to the surface of the oxide film. Taking advantage of the characteristic that zinc yellow pigment gradually releases chromate ions after being exposed to moisture and forms color indicator bands inside the coating and at the edge of defects, a color difference indicator system that can self-heal over service time is constructed to obtain a coating test piece as the object of subsequent testing.

[0006] The coated test specimens were placed in a coastal atmospheric exposure environment, and samples were taken and tested at preset time intervals. The surface parameters of the coating on the exposed test specimens were measured, and the change curves of each parameter on the surface of the coating with the exposure time were continuously recorded. Simultaneously, a dynamic correction coefficient was calculated in real time based on the degree of hydrolysis of zinc yellow pigment. The dynamic correction coefficient was updated at a uniform ratio to the gloss loss rate threshold, color difference threshold, contact angle threshold, defect area threshold, and corrosion current density threshold of the coating surface parameters in the previous cycle, so as to obtain a set of thresholds that match the aging degree of the sample.

[0007] Electron microscopy was used to observe the exposed test specimens sampled at the same time. The micro-defect region was obtained by image segmentation algorithm, the defect area ratio was calculated and the defect boundary complexity was quantitatively characterized. The polarization curve of the same batch of exposed test specimens was tested by electrochemical workstation, and the corrosion current density was obtained by Tafel extrapolation method.

[0008] The five measurement results of the coating surface parameters measured in this cycle—gloss loss rate, color difference value, water droplet contact angle, defect area ratio, and corrosion current density—are compared with the dynamically updated threshold set. When all five measurement results are higher than the corresponding threshold in the threshold set, the coating complete failure judgment information is output.

[0009] Compared with the prior art, the beneficial effects of this invention are as follows:

[0010] This invention provides a method for detecting surface defects in epoxy-coated aluminum alloy plates. The method includes: anodizing the aluminum alloy plate substrate to form an oxide film, and then coating the entire surface of the oxide film with an epoxy resin coating containing zinc yellow pigment; utilizing the characteristic that zinc yellow pigment gradually releases chromate ions upon exposure to moisture and forms color indicator bands within the coating and at defect edges, constructing a self-healing color difference indicator system over service time to obtain a coating test piece for subsequent testing; placing the coating test piece in a coastal atmospheric exposure environment, sampling and testing the exposed test piece at preset time intervals, measuring the coating surface parameters of the exposed test piece, and continuously recording the change curves of each parameter with exposure time; simultaneously calculating a dynamic correction coefficient in real time based on the degree of hydrolysis of the zinc yellow pigment, wherein the dynamic correction coefficient is calculated at a uniform multiplier. The previous cycle's coating surface parameters, including gloss loss rate threshold, color difference threshold, contact angle threshold, defect area threshold, and corrosion current density threshold, were updated to obtain a threshold set that matched the sample's aging degree. Electron microscopy was used to observe exposed test specimens sampled concurrently. Image segmentation algorithms were employed to obtain microscopic defect regions, calculate the defect area ratio, and quantitatively characterize the defect boundary complexity. An electrochemical workstation was used to perform polarization curve testing on the same batch of exposed test specimens, and the corrosion current density was obtained using the Tafel extrapolation method. The five measurement results of the coating surface parameters measured in this cycle—gloss loss rate, color difference value, water droplet contact angle, defect area ratio, and corrosion current density—were compared with the dynamically updated threshold set. When all five measurement results exceeded the corresponding threshold in the threshold set, a complete coating failure determination was output. This method enables real-time tracking of zinc yellow pigment hydrolysis behavior, automatic correction of multiple thresholds, and integration of optical, morphological, and electrochemical information to accurately assess the failure risk of epoxy-coated aluminum alloy plates under coastal service conditions. Attached Figure Description

[0011] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. Some specific embodiments of the invention will be described in detail below with reference to the accompanying drawings in an exemplary and non-limiting manner. The same reference numerals in the drawings designate the same or similar parts or components. It should be understood by those skilled in the art that these drawings are not necessarily drawn to scale. In the drawings:

[0012] Figure 1 This is a schematic flowchart of a method for detecting surface defects in epoxy-coated aluminum alloy plates according to an embodiment of the present invention.

[0013] Figure 2These are comparative images of the microscopic morphology of surface defects of epoxy-coated aluminum alloy plates after different exposure times according to embodiments of the present invention; where a corresponds to 7 years of exposure; b corresponds to 12 years of exposure; and c corresponds to 20 years of exposure. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of this embodiment will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] See Figures 1-2 This embodiment provides a method for detecting surface defects in epoxy-coated aluminum alloy plates, comprising the following steps:

[0016] S101. The aluminum alloy plate substrate is anodized to form an oxide film, and an epoxy resin coating containing zinc yellow pigment is applied to the surface of the oxide film. Taking advantage of the characteristic that zinc yellow pigment gradually releases chromate ions after being exposed to moisture and forms color indicator bands inside the coating and at the edge of defects, a color difference indicator system that can self-repair over service time is constructed to obtain a coating test piece as the object of subsequent testing.

[0017] S102. Place the coating test specimen in a coastal atmospheric exposure environment, and sample and test the exposed test specimen at preset equal time intervals. Measure the coating surface parameters of the exposed test specimen respectively, and continuously record the change curves of each parameter in the coating surface parameters with the exposure time. Simultaneously, calculate the dynamic correction coefficient in real time based on the degree of hydrolysis of zinc yellow pigment. The dynamic correction coefficient updates the gloss loss rate threshold, color difference threshold, contact angle threshold, defect area threshold, and corrosion current density threshold in the coating surface parameters of the previous cycle at a uniform ratio to obtain a threshold set that matches the aging degree of the sample.

[0018] S103. Electron microscopy was used to observe the exposed test pieces sampled at the same time. The micro-defect area was obtained by image segmentation algorithm, the defect area ratio was calculated and the defect boundary complexity was quantitatively characterized. The polarization curve of the same batch of exposed test pieces was tested by electrochemical workstation, and the corrosion current density was obtained by Tafel extrapolation method.

[0019] S104. Compare the five measurement results of the coating surface parameters measured in this cycle, namely gloss loss rate, color difference value, water droplet contact angle, defect area ratio and corrosion current density, with the threshold set obtained by dynamic update. When all five measurement results are higher than the corresponding threshold in the threshold set, output the coating complete failure judgment information.

[0020] It should be noted that in this embodiment, the epoxy coating doped with zinc yellow pigment releases chromate ions that form color-changing bands in the damaged areas of the coating, achieving early visual indication of defects and improving the intuitiveness and responsiveness of defect identification. A dynamic correction coefficient is calculated using the hydrolytic behavior of zinc yellow pigment, and various judgment thresholds are adaptively adjusted based on the sample aging degree, allowing the detection standards to be automatically updated over time, avoiding misjudgments caused by using fixed static thresholds. By comprehensively comparing and judging five types of data—gloss loss rate, color difference value, water droplet contact angle, defect area ratio, and corrosion current density—the robustness and accuracy of failure judgment are enhanced, reducing the uncertainty of judgment based on a single indicator.

[0021] It should be noted that in this embodiment, electron microscopy combined with image segmentation algorithms is used to automatically identify microscopic defect regions without relying on human experience, improving the accuracy and consistency of defect identification. This is particularly effective for fine cracks, corrosion points, or unevenly distributed spalling areas, enabling high-resolution detection and objective quantitative analysis. By calculating the defect area ratio, the proportion of the damaged area on the entire sample surface is quantified, providing a numerical indicator of the coating degradation degree and a basis for subsequent lifetime prediction, graded early warning, and coating performance evaluation. Analyzing the defect boundary morphology (such as fractal dimension and other complexity indicators) helps to distinguish different types of defects (such as pitting, line cracks, and spalling) and trace their formation mechanisms, improving adaptability to complex defects. Polarization curve testing is performed using an electrochemical workstation, and the corrosion current density is obtained using Tafel extrapolation, reflecting the material's corrosion sensitivity under service conditions. This compensates for the limitations of optical and morphological indicators, providing intuitive quantitative support for the material's electrochemical behavior. By combining morphological changes with electrochemical corrosion data, multi-dimensional consistency verification is achieved, effectively improving the reliability and robustness of failure determination.

[0022] It should be noted that in this embodiment, the anodic oxide film refers to a dense oxide layer formed on the aluminum alloy surface through electrochemical anodizing, used to improve corrosion resistance and adhesion. Zinc yellow pigment is a chromate-containing anti-corrosion pigment that can slowly release chromate ions in humid environments, providing both corrosion inhibition and color indication functions. The color difference indication system refers to the visible color difference area formed by the reaction of zinc yellow pigment exudate products with the corroded area, used to determine localized coating failure. Coating surface parameters are quantifiable parameters used to characterize changes in coating performance, including gloss loss rate, color difference, and water droplet contact angle. The dynamic correction coefficient is a multiplier calculated in real-time based on the degree of hydrolysis release of zinc yellow pigment, used to update the aging threshold to adapt to the accuracy requirements of different exposure stages. The Tafel extrapolation method is a standard method used in electrochemical testing to calculate corrosion current density, obtained by extrapolating the logarithmic linear region of the polarization curve. Corrosion current density is a parameter characterizing the degree of electrochemical corrosion of the coating; a higher value indicates more severe corrosion.

[0023] Preferably, the gloss loss rate is measured using a standard gloss meter. At least twenty measurement points on the test piece surface are tested under a uniform incident angle. After excluding edge effects, the arithmetic mean is taken as the gloss loss rate for this period. The threshold for the gloss loss rate is determined by fitting an exponential decay curve to the early exposure gloss decay curve. By setting a uniform incident angle and sampling at multiple points (no fewer than twenty) on the test piece surface, while excluding outliers in the edge region, the random errors from single-point testing and the influence of local surface inhomogeneities can be reduced. This results in a more accurate and statistically representative average gloss loss rate, ensuring the repeatability and stability of the evaluation results. Taking the arithmetic mean as the gloss loss rate for this period after excluding edge effects can eliminate interference areas such as mechanical shearing, inconsistent thickness, or surface inhomogeneity that may exist at the coating edge, avoiding deviations in the overall measurement value due to abnormal optical response in the edge region. This ensures that the data accurately reflects the service status of the coating's central region from the source. The threshold for gloss loss rate is determined by fitting the gloss decay curve in the early stage of exposure with an exponential decay method. This not only allows for setting criteria based on the decay law of the sample itself, but also adapts to the differences between samples from different batches and under different exposure environments, thereby improving the scientific nature and pertinence of the judgment threshold and providing reliable support for subsequent failure warnings.

[0024] Preferably, the water droplet contact angle in the coating surface parameters of the exposed test specimen is determined using a rotating droplet analyzer. Thirty measurement points are randomly selected for each sampled test specimen, and the Kolmogorov-Smirnov test is used to ensure that the measurement data are normally distributed. Under the condition that the significance level P>0.05 is met, the surface free energy of the coating is calculated using the following formula: In the formula For the dispersion component, It is a polar component; when the surface free energy of the coating is... When the contact angle parameter exhibits a monotonically increasing trend with exposure time and its first derivative exceeds the threshold derivative calculated from the zinc yellow hydrolysis yield, the contact angle parameter is set to an out-of-limit state.

[0025] It should be noted that traditional contact angle measurements typically only take a small number of fixed-point data. In this embodiment, thirty measurement points are randomly selected from each sample, and the normal distribution is verified by the Kolmogorov-Smirnov test. Skewed or outlier values ​​are actively removed to ensure sample representativeness and data robustness, reducing the interference of accidental factors such as coating roughness and local contamination on the results. In addition, the use of a rotating droplet analyzer allows for continuous and higher-precision multi-point testing on the same platform, avoiding droplet volume fluctuations and optical path errors caused by manual sample changes. Combined with a random point selection strategy, uniform sampling within a large sample area is achieved. Thus, this embodiment breaks through the limitations of conventional static droplet single-point measurement, making the contact angle results more sensitive to progressive defects such as large-area coating gloss loss and cracking. Moreover, this embodiment does not stop at the static comparison of the absolute value of surface free energy, but calculates the surface free energy of the coating in real time. The hydrophilicity transition point of the coating is dynamically determined by comparing the monotonically increasing rate of exposure time with a threshold derived from the hydrolysis yield of zinc yellow pigment. This allows for early identification of critical points where the pigment's corrosion inhibition ability is about to deteriorate. When the surface free energy growth rate exceeds the threshold, the contact angle index is automatically marked as exceeding the limit, enhancing the overall judgment sensitivity. The Kolmogorov-Smirnov test (KS test) is a nonparametric statistical test used to determine whether a set of sample data conforms to a known distribution.

[0026] Preferably, when using an image segmentation algorithm to obtain the microscopic defect region, calculate the defect area ratio, and quantitatively characterize the defect boundary complexity, the process includes: performing binarization processing on the image obtained by electron microscopy scanning, determining the grayscale threshold using the Otsu adaptive algorithm to achieve optimal separation between defective and non-defective pixels under the condition of maximizing global variance; statistically analyzing the ratio of defective pixels to total pixels to obtain the defect area ratio; performing coordinate sampling on the boundary of each defect connected region; calculating the fractal dimension D using an image fractal analysis algorithm; and using the fractal dimension D as the sole metric for boundary complexity; and utilizing the natural staining effect of zinc yellow pigment continuously releasing chromate ions at the edge of the defect depression and forming a visible chromate halo to provide a color reference for defect boundary extraction. It should be noted that in this embodiment, the image obtained by electron microscopy is binarized, and the grayscale threshold is determined by the Otsu adaptive algorithm. This ensures optimal separation between defective and non-defective pixels under the condition of maximizing global variance, avoiding the problem of unstable segmentation performance under different brightness and background textures in the fixed threshold method. This enhances the robustness to complex morphological surfaces and makes defect recognition stable in both high-contrast and low-contrast images. The ratio of defective pixels to total pixels is statistically analyzed to obtain the defect area ratio. The boundaries of each defective connected region are sampled by coordinates, and the fractal dimension D is calculated using an image fractal analysis algorithm. The fractal dimension D is used as the sole metric for boundary complexity, allowing the previously subjective descriptions of edge roughness, contour complexity, etc., to be uniformly represented in digital form. This helps to further determine whether the defect is erosion-type, diffusion-type, or tear-type, providing more diagnostically valuable morphological features. By utilizing the natural staining effect of zinc yellow pigment continuously releasing chromate ions at the edge of the defect depression and forming a visible chromate halo, a color reference is provided for the extraction of the defect boundary. This enables the processing of images with blurred edges due to indistinct gray-scale transitions or complex surface undulations, thereby improving the physical correspondence and interpretability of the segmented contour.

[0027] Preferably, the fractal dimension D is obtained by linearly fitting the perimeter L of the defect region boundary with the corresponding area A using a double logarithmic scale. The mathematical model for the fitting is: logL = k·logA + b, D = 2(k + 1). For the same defect connected region, if the growth rate ΔD / Δt of the fractal dimension obtained in two adjacent exposure samplings exceeds the critical growth curve obtained by mapping the zinc yellow pigment hydrolysis rate r through an empirical function, then the defect is determined to have entered a rapid runaway stage from the slow expansion stage. k is the slope of the fitted line, representing the degree of response of the defect boundary complexity to changes in area; b is the intercept of the fitted line, reflecting the initial value of boundary growth at the initial scale. It should be noted that in this embodiment, by establishing a calculation model for the fractal dimension through the double logarithmic fitting relationship between the perimeter and area of ​​the defect region boundary, the complexity of the defect boundary can be accurately quantified by mathematical formulas, overcoming the limitations of relying solely on subjective judgments of area or shape. Furthermore, this embodiment not only focuses on the static value of the fractal dimension but also provides the growth rate of the fractal dimension between consecutive samples as a dynamic criterion, and further compares it with the critical growth rate curve constructed from the hydrolysis rate of zinc yellow pigment. When the complexity growth rate of the defect boundary exceeds this critical value, it is determined that the defect has entered a rapid runaway stage, thereby triggering an early warning. Compared with approaches that only focus on static geometric parameters or rely on a single corrosion index, this embodiment achieves a deep coupling between image morphology evolution and pigment chemical reaction behavior, possessing stronger predictive and forward-looking capabilities.

[0028] Preferably, the specific steps of using an electrochemical workstation to perform polarization curve testing on the sampled exposed test pieces from the same batch and obtaining the corrosion current density using the Tafel extrapolation method include: connecting the exposed test piece to a standard electrochemical testing system, which includes a three-electrode configuration of a working electrode, a reference electrode, and an auxiliary electrode, wherein the working electrode is the coated test piece; placing the test piece in a set electrolytic medium before testing, and starting the test after stabilization treatment; applying a continuously changing potential to the working electrode through the electrochemical workstation and simultaneously recording the response current to obtain the potential-current relationship curve data; selecting a stable region from the obtained potential-current relationship curve and analyzing the anodic and cathodic trends of the stable region; determining the location of the corrosion potential based on the trend of the curve slope change; and deriving the corresponding corrosion current density. It should be noted that in this embodiment, by performing polarization curve tests on the same batch of exposed test pieces under a three-electrode system and performing slope analysis on the linear segments of the anode and cathode, the corrosion potential is accurately located and the corrosion current density is derived from it. By using continuous potential scanning and whole-segment trend fitting, instantaneous noise and polarization fluctuations can be filtered out, and more stable and repeatable corrosion current density results can be given. The location of the corrosion potential is determined based on the curve slope change trend, and the corresponding corrosion current density is derived, thereby improving the accuracy of coating corrosion state quantification.

[0029] Preferably, the specific steps for determining the location of the corrosion potential based on the trend of the curve slope and deriving the corresponding corrosion current density include: organizing and preprocessing the potential and current response data collected during the electrochemical test to remove interference segments and unstable segments; identifying the linear trend regions of the anodic and cathodic regions in the preprocessed potential-current relationship curve; performing fitting analysis on the potential and logarithmic current data of the two regions respectively; determining the corrosion potential reflecting the electrochemical stable state of the test piece by the intersection of the fitted curves; and extrapolating the corrosion current density at the corrosion potential location based on the fitted trend. It should be noted that in this embodiment, the polarization curve data is structured through the linear trend identification of the anodic and cathodic regions and the logarithmic current-potential data fitting strategy, significantly improving the scientific rigor and reliability of corrosion current density extraction. It no longer relies on manual point selection or single-interval slopes, but systematically identifies the most representative linear response region in the curve. By fitting the slopes of the anodic and cathodic regions respectively and determining their intersection locations, the corrosion potential is accurately defined and the corrosion current density is extrapolated, thereby avoiding the problems of noise interference and strong subjectivity in data selection in the traditional Tafel method. In addition, this embodiment also removes unstable and interfering segments through data preprocessing, which effectively enhances the consistency and representativeness of the fitting interval and improves the robustness of corrosion rate judgment.

[0030] Preferably, the specific steps for organizing and preprocessing the potential and current response data collected during the electrochemical testing process to remove interference and unstable regions include: completely exporting the raw test data output from the electrochemical workstation and constructing a data sequence according to the correspondence between potential and current; performing preliminary screening on the constructed data sequence to eliminate obvious abrupt changes caused by electrode polarization instability or contact noise at the beginning of the test, and identifying and removing nonlinear convergence or distorted data caused by the end of the scan; and smoothing the remaining data region to remove local current fluctuation interference. It should be noted that in this embodiment, by implementing systematic preprocessing of the raw polarization curve data, higher accuracy and repeatability are achieved in the electrochemical analysis. Specifically, the complete data is first exported and a sequence is constructed according to the potential and current order to ensure that no sampling information is lost in subsequent processing. Then, a regular screening process is used to automatically remove unstable polarization of the starting electrode, sudden changes in contact noise, and nonlinear distortion sections at the end of the scan, effectively avoiding subjective errors caused by manual truncation. Finally, a smoothing operation is performed on the remaining sequence to eliminate local current jitter and make the data show a regular monotonic trend, which is convenient for accurate fitting of the linear regions of the anode and cathode.

[0031] Preferably, the specific steps for synchronously calculating the dynamic correction coefficient in real time based on the degree of hydrolysis of zinc yellow pigment include: in each sampling cycle, collecting hydrolysis-related signals of zinc yellow pigment in real time. These signals are obtained by detecting changes in the absorbance of chromate in the coating exudate or reading the color difference value of the coating color indicator band, and recording the difference between the hydrolysis-related signals and the reference signal of the previous cycle; inputting the collected hydrolysis-related signals into a pre-established mapping model, which converts the hydrolysis signals into a hydrolysis process index to characterize the relative degree of zinc yellow pigment hydrolysis in the current cycle; and retrieving the corresponding dynamic correction coefficient based on the hydrolysis process index. The dynamic correction coefficient maintains a monotonic relationship with the degree of hydrolysis, reflecting the magnitude of the impact of pigment-released chromate on the attenuation of coating protective performance. It should be noted that in this embodiment, by tracking the amount of chromate exudation in real time through absorbance or color difference, the pigment corrosion inhibition activity can be continuously quantified without damaging the coating, improving monitoring frequency and on-site applicability. In addition, in this embodiment, the hydrolysis signal is converted into a hydrolysis process index to characterize the relative degree of hydrolysis of zinc yellow pigment in the current period. The corresponding dynamic correction coefficient is retrieved according to the hydrolysis process index. The dynamic correction coefficient maintains a monotonic relationship with the degree of hydrolysis to reflect the influence of pigment slow-release chromate on the attenuation of coating protective performance, thereby avoiding early false alarms or late omissions due to chemical slow-release effect of fixed threshold.

[0032] Preferably, the specific steps for updating the threshold values ​​of gloss loss rate, color difference, contact angle, defect area, and corrosion current density in the coating surface parameters of the previous cycle with the dynamic correction coefficient at a uniform multiple to obtain a threshold set matching the aging degree of the sample include: after obtaining the dynamic correction coefficient for the current cycle, calling the basic threshold set stored in the previous cycle, which contains the original judgment reference values ​​for the five coating surface parameters; judging the change trend of the degree of zinc yellow pigment hydrolysis in the current cycle compared to the previous cycle, and confirming whether the direction of the dynamic correction coefficient is to relax or tighten; applying the dynamic correction coefficient to the basic threshold set to uniformly adjust the threshold values ​​of the five coating surface parameters, with the adjustment rules applied to the threshold values ​​of gloss loss rate, color difference, contact angle, defect area, and corrosion current density according to the set weight factors; performing boundary verification on the adjusted threshold values ​​to ensure that the updated results are not lower than the set minimum reference safety threshold, nor exceed the allowed maximum stable threshold range; updating the verified adjustment results to the effective threshold set for the current cycle, and using it for failure comparison of the measurement results in this cycle, while storing this threshold set as the basic threshold set for the next cycle. It should be noted that in this embodiment, the basic threshold of the previous cycle is first called, and then the threshold should be tightened or loosened automatically based on the changes in pigment hydrolysis. The five parameters are adjusted uniformly according to the preset weight to ensure that each index remains relatively balanced under the same decay scale. Then, upper and lower boundaries are set to prevent the threshold from being infinitely amplified or reduced due to short-term fluctuations, and to maintain the safe and effective range of the judgment standard. Finally, the new threshold set that has been verified is immediately put into the comparison of the current cycle and used as the benchmark for the next cycle to realize closed-loop recursive update. This achieves multi-parameter synchronous, adaptive, and controlled threshold management, which can continuously track the chemical self-healing decay and physical damage accumulation in the coating aging process and improve the accuracy and timeliness of failure warning.

[0033] Furthermore, the specific steps for calculating the dynamic correction coefficient in real time include: measuring the concentration of chromate ions generated by the hydrolysis of zinc yellow pigment on the surface of the coated test piece in real time, and establishing a chromate concentration curve characterizing the degree of hydrolysis of zinc yellow pigment based on the rate of change of chromate ion concentration at different exposure stages; calculating the real-time slope of the chromate concentration curve using a sliding window time series analysis method to determine the dynamic trend of the hydrolysis rate of zinc yellow pigment with exposure time, and identifying the trend inflection point based on a preset slope change threshold; automatically adjusting the update ratio of the dynamic correction coefficient according to the trend inflection point identification result, wherein the correction coefficient ratio is increased when the hydrolysis rate accelerates beyond the threshold, and decreased when the hydrolysis rate slows down below the threshold. It should be noted that in this embodiment, by monitoring the chromate ion release rate in real time and automatically adjusting the dynamic correction coefficient ratio based on the trend inflection point, the problem of threshold update lag caused by the nonlinear change of the zinc yellow pigment hydrolysis rate is effectively solved, improving the threshold adaptation accuracy and the real-time performance and reliability of coating failure determination.

[0034] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method of detecting surface defects of an aluminum alloy panel with an epoxy coating, characterized by, The method comprises the following steps: An anodic oxidation treatment is performed on an aluminum alloy plate base to form an oxide film, and an epoxy resin coating doped with zinc yellow pigment is applied on the surface of the oxide film; by using the characteristic of the zinc yellow pigment gradually releasing chromate ions after being wetted and forming a color indicating band inside the coating and at the edge of defects, a color difference indicating system capable of self-repairing with service time is constructed to obtain a coating test piece as a subsequent detection object; The coating test piece is placed in a coastal atmospheric exposure environment, and at predetermined time intervals, the exposed test piece is sampled and detected, the coating surface parameters of the exposed test piece are measured respectively, and the change curves of each parameter in the coating surface parameters with exposure time are continuously recorded; a dynamic correction coefficient is calculated in real time according to the hydrolysis degree of the zinc yellow pigment, the dynamic correction coefficient updates the light loss rate threshold, the color difference threshold, the contact angle threshold, the defect area threshold and the corrosion current density threshold in the coating surface parameters in the previous period by a unified ratio to obtain a threshold set matched with the aging degree of the sample; The exposed test pieces sampled at the same period are observed by scanning electron microscopy, the microscopic defect area is obtained by using an image segmentation algorithm, the defect area ratio is calculated and the defect boundary complexity is quantitatively characterized, and the polarization curve test of the sampled exposed test pieces of the same batch is carried out by using an electrochemical workstation, and the corrosion current density is obtained by Tafel extrapolation method; The five measurement results of the light loss rate, the color difference value, the water droplet contact angle, the defect area ratio and the corrosion current density in the coating surface parameters measured in the current period are compared with the threshold set obtained by dynamic updating, and when all the five measurement results are higher than the corresponding threshold in the threshold set, coating complete failure determination information is outputted; The specific steps of the real-time calculation of the dynamic correction coefficient according to the hydrolysis degree of the zinc yellow pigment include: In each sampling period, the zinc yellow pigment hydrolysis related signal is collected in real time, the hydrolysis related signal is obtained by detecting the absorbance change of chromate in the coating exudate or reading the color difference value of the color indicating band of the coating, and the difference between the hydrolysis related signal and the reference signal in the previous period is recorded; The collected hydrolysis related signal is input into a pre-established mapping model, the mapping model converts the hydrolysis signal into a hydrolysis process index for representing the relative degree of zinc yellow pigment hydrolysis in the current period; the corresponding dynamic correction coefficient is retrieved according to the hydrolysis process index, the dynamic correction coefficient and the hydrolysis degree maintain a monotonic relationship, which reflects the influence amplitude of the pigment releasing chromate on the attenuation of the coating protection performance.

2. The method of claim 1, wherein the method is characterized by: The measurement of the light loss rate uses a standard gloss meter to detect at least twenty measurement points on the surface of the test piece under the condition of a unified incident angle, and the arithmetic mean value after excluding the edge effect is taken as the light loss rate in the current period, and the threshold of the light loss rate is obtained by exponential decay fitting of the gloss decay curve in the early exposure period.

3. The method of claim 1, wherein the method is characterized by: The water drop contact angle in the coating surface parameter of the exposed test piece is determined by a rotating drop analyzer, thirty measurement points are randomly selected for each sample test piece for testing, and the Kolmogorov-Smirnov test is used to ensure that the measurement data is normally distributed; under the condition of meeting the significance level P>0.05, the coating surface free energy is calculated by the following formula: ; In the formula is the dispersion component, is the polar component; when the coating surface free energy monotonically increases with exposure time and its first derivative is higher than the threshold derivative calculated by the zinc yellow hydrolysis rate, the contact angle parameter is set to an over-limit state.

4. The method of claim 1, wherein the method is characterized by: The image segmentation algorithm is used to obtain the micro defect area, calculate the defect area ratio and quantitatively characterize the defect boundary complexity, including: performing binaryzation processing on the image obtained by scanning the electron microscope, and determining the gray threshold value by Otsu adaptive algorithm, so that the defect pixels and non-defect pixels are optimally separated under the condition of global variance maximization; the proportion of defect pixels to total pixels is counted to obtain the defect area ratio, the coordinates of the boundary of each defect connected domain are sampled, and the fractal dimension D is calculated by using the image fractal analysis algorithm, and the fractal dimension D is used as the only measurement index of the boundary complexity; the natural dyeing effect of zinc yellow pigment continuously releasing chromate and forming a visible chromate halo at the edge of the defect depression is used as a color reference for defect boundary extraction.

5. The method of claim 4, wherein the method is characterized by: The numerical value of the fractal dimension D is obtained by linear fitting of the double logarithmic scale of the defect area boundary perimeter L and the corresponding area A, and the mathematical model of the fitting is: logL=k·logA+b, D=2(k + 1); for the same defect connected domain, if the growth rate ΔD / Δt of the fractal dimension obtained in the adjacent two exposure samplings exceeds the critical growth rate curve obtained by mapping the hydrolysis rate r of zinc yellow pigment by an empirical function, it is determined that the defect has entered the rapid out-of-control stage from the slow expansion stage; k is the slope of the fitting straight line, indicating the response degree of the defect boundary complexity to the area change; b is the intercept of the fitting straight line, reflecting the initial value of the boundary growth at the starting scale.

6. The method of claim 1, wherein the method is characterized by: The specific steps of the polarization curve test of the same batch of exposure test pieces by the electrochemical workstation and the Tafel extrapolation method to obtain the corrosion current density include: connecting the exposure test piece to a standard electrochemical test system, the standard electrochemical test system including a three-electrode configuration of a working electrode, a reference electrode and an auxiliary electrode, wherein the working electrode is the coating test piece; placing the test piece in a set electrolyte before testing, and starting testing after stabilization treatment; applying a continuously changing potential to the working electrode by the electrochemical workstation, and synchronously recording the response current to obtain the curve data of the potential-current relationship, selecting the stable region from the obtained potential-current relationship curve, and analyzing the anodic and cathodic trend parts of the stable region, determining the position of the corrosion potential according to the change trend of the curve slope, and deducing the corresponding corrosion current density.

7. The method of claim 6, wherein the method is characterized by: The specific steps of determining the position of the corrosion potential according to the change trend of the curve slope and deducing the corresponding corrosion current density include: arranging and preprocessing the potential and current response data collected during the electrochemical test to remove interference sections and non-stable sections; identifying the linear trend regions of the anode and cathode regions in the potential-current relationship curve after preprocessing, fitting and analyzing the potential and logarithmic current data of the two regions respectively, determining the corrosion potential reflecting the electrochemical stable state of the test piece through the intersection position of the fitting curve, and extrapolating the corrosion current density at the corrosion potential position according to the fitting trend.

8. The method of claim 7, wherein the method is characterized by: The specific steps of collating and preprocessing the potential and current response data collected in the electrochemical test process include: exporting the original test data output by the electrochemical workstation in full, and constructing a data sequence according to the correspondence between the potential and the current; preliminarily screening the constructed data sequence, eliminating the obvious mutation section caused by electrode polarization instability or contact noise in the beginning stage of the test, and identifying and removing the nonlinear convergence or distortion data caused by the end of scanning at the tail; smoothing the remaining data area to remove local current fluctuation interference.

9. The method of claim 1, wherein the method is used to detect surface defects of an epoxy-coated aluminum alloy panel. The specific steps of updating the light loss rate threshold, the color difference threshold, the contact angle threshold, the defect area threshold and the corrosion current density threshold in the coating surface parameters of the last period by the dynamic correction coefficient according to a uniform ratio to obtain a threshold set matched with the aging degree of the sample include: After obtaining the dynamic correction coefficient of the current period, the basic threshold set stored in the last period is called, and the basic threshold set includes the original judgment reference values for the five coating surface parameters; The change trend of the hydrolysis degree of the zinc yellow pigment in the current period compared with the last period is judged to confirm the action direction of the dynamic correction coefficient as relaxation or tightening; the dynamic correction coefficient is applied to the basic threshold set to uniformly adjust the thresholds of the five coating surface parameters, and the adjustment rule is applied to the light loss rate, the color difference, the contact angle, the defect area and the corrosion current density according to the set weight factor; The adjusted thresholds are verified at the boundary to ensure that the updated results are not lower than the set minimum reference safety threshold and are not more than the allowed maximum stable threshold range; the adjustment results that pass the verification are updated as the effective threshold set of the current period and are used for failure comparison of the measurement results in the current period, and the threshold set is stored as the basic threshold set of the next period.

Images