Cover plate defect discrimination method and system based on cleaning tempered screen printing data fusion

Abstract

The application relates to the technical field of cover plate defect identification, in particular to a cover plate defect identification method and system based on cleaning tempered silk screen data fusion. The method comprises the following steps: obtaining an irregular curved surface cover plate information set, analyzing hidden ion water mark residual information caused by the limitation of the curved surface structure and chemical tempering based on the irregular curved surface cover plate information set, and obtaining a residual information set; based on the residual information set, analyzing the secondary ion exchange effect of ink low-melting-point glass powder and the chemical tempering layer during silk screen printing and the surface stress distribution change caused by the secondary ion exchange effect, and obtaining a defect correlation information set; and generating and outputting cover plate defect identification information according to the defect correlation information set. The method effectively solves the problem that in the prior art, there is a lack of effective pre-identification and cause tracing means for the hidden defects caused by the coupling of multiple process parameters in the manufacturing process of the irregular curved surface cover plate.

Description

Technical Field

[0001] This application relates to the field of cover plate defect identification technology, and in particular to a cover plate defect identification method and system based on the fusion of cleaning, tempering, and screen printing data. Background Technology

[0002] With the rapid development of electronic devices towards full-screen, curved screen and foldable designs, irregularly shaped curved cover plates have been widely used in terminal products such as smartphones and tablets due to their excellent grip and aesthetics. Among these processes, cleaning, chemical tempering and screen printing have a decisive impact on the final appearance quality of the cover plate. In the existing production system, the detection of defects such as dirt, watermarks, pinholes, uneven ink and rainbow patterns on the cover plate surface mainly relies on manual visual inspection or automated optical inspection (AOI) equipment based on optical imaging principles.

[0003] However, in the existing technology, there is a lack of effective means to identify and trace the causes of hidden defects caused by the coupling of multiple process parameters during the manufacturing of irregular curved surface covers. In particular, those defects that are formed during the cleaning and tempering stages but only become macroscopically abnormal after screen printing and sintering are often difficult to be accurately captured and identified in the early stages, resulting in a high risk of defective products flowing out and difficulty in locating the specific failure point at the root. Summary of the Invention

[0004] This application provides a method and system for identifying cover plate defects based on the fusion of cleaned and tempered screen printing data, in order to solve the above-mentioned problems.

[0005] In a first aspect, this application provides a method for judging cover plate defects based on the fusion of screen printing data from cleaning and tempering processes. The method includes: acquiring an information set of irregularly shaped curved cover plates; analyzing the residual information of invisible ion water marks caused by the limitations of the curved surface structure and chemical tempering based on the information set of irregularly shaped curved cover plates to obtain a residual information set; analyzing the secondary ion exchange effect between the low-melting-point glass powder of the ink and the chemical tempering layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect, based on the residual information set to obtain a defect association information set; and generating and outputting cover plate defect judgment information according to the defect association information set.

[0006] The above technical solution first relies on the information set of irregular curved surface cover plates to accurately extract the residual information set of invisible ion water marks, solving the problem of water marks being difficult to identify due to curved surface structure and chemical tempering. Then, based on the residual information set, it deeply analyzes the secondary ion exchange effect of ink low melting point glass powder and chemical tempering layer and the changes in surface stress distribution, establishing a correlation link from residual characteristics to defect causes. This breaks through the limitation of existing defect identification that only focuses on surface appearance and ignores internal stress and ion exchange causes. Finally, it generates accurate identification information based on the defect correlation information set, realizing integrated identification from method data to defect results. This significantly improves the accuracy and comprehensiveness of defect identification of irregular curved surface cover plates, effectively reduces the outflow of defective products, and provides data support for method optimization, improving the overall quality and production efficiency of cover plate processing.

[0007] Optionally, based on the information set of the irregular curved cover plate, the analysis of the residual information of invisible ion water marks caused by the curvature structure and chemical tempering to obtain the residual information set includes: the information set of the irregular curved cover plate includes a cleaning information set, a tempering information set, and a screen printing information set; based on the cleaning information set and combined with the tempering information set, the analysis of the distribution of cleaning fluid residue area caused by the change of curvature of the irregular curved cover plate during the cleaning process, and the mismatch information between the ion exchange depth and the residue area during the chemical tempering process, to obtain preliminary residual information; based on the preliminary residual information and combined with the screen printing information set, the analysis of the coverage uniformity and masking effect of ink in the preliminary residue area during the screen printing process, to obtain the residual information set.

[0008] Optionally, the process of constructing the preliminary residual information includes: based on the cleaning information set, analyzing the liquid film thickness distribution formed in the concave and convex regions of the irregular curved cover plate due to the difference in surface tension of the cleaning fluid and the different drying rates, to determine the cleaning fluid residual information; based on the tempering information set, analyzing the difference in diffusion rate of ion exchange depth in the planar region and the curvature change region during chemical tempering due to the stress gradient, to obtain ion exchange distribution information; comparing the cleaning fluid residual information with the ion exchange distribution information, and extracting the spatially overlapping region where the ion exchange depth is lower than the depth of the residual region, as the preliminary residual information.

[0009] Optionally, the process of constructing the ion exchange distribution information includes: based on the tempering information set, analyzing the stress distribution differences caused by the different curvatures of the irregular curved surface cover plate in the planar region and the curvature variation region, to obtain tempering stress distribution information; based on the tempering stress distribution information, analyzing the modulation direction and modulation amplitude of the stress gradient in each region on the ion diffusion activation energy, to obtain the ion exchange distribution information.

[0010] Optionally, the step of analyzing the uniformity of ink coverage and masking effect in the initial residual area during screen printing, based on the preliminary residual information and the screen printing information set, to obtain the residual information set includes: based on the preliminary residual information, analyzing the local micro-separation state caused by the difference in shrinkage stress during the curing and shrinkage process of the ink to obtain coverage uniformity information; based on the coverage uniformity information and the screen printing information set, analyzing the influence of the stress difference between the residual area and the non-residual area of ​​the cured ink layer on the optical transmission characteristics to obtain masking effect information; and integrating the coverage uniformity information and the masking effect information to construct the residual information set.

[0011] Optionally, the step of analyzing the secondary ion exchange effect between the low-melting-point glass powder of the ink and the chemically tempered layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect, based on the residual information set, to obtain a defect association information set includes: based on the residual information set, analyzing the interfacial contact gap caused by the local micro-separation state of the low-melting-point glass powder of the ink in the residual area, and the contact characteristics formed by stress differences in the non-residual area, to obtain interfacial contact characteristic information; based on the interfacial contact characteristic information, analyzing the differences in ion exchange activation energy caused by different contact states between the ink layer and the chemically tempered layer in each region during sintering, to obtain secondary ion exchange information; based on the secondary ion exchange information, analyzing the changes in surface layer ion concentration in each region caused by differences in potassium and sodium ion exchange, and the resulting redistribution of local compressive stress, to obtain the defect association information set.

[0012] Optionally, the process of constructing the secondary ion exchange information includes: based on the interface contact characteristic information, analyzing the thermal expansion behavior of residual gas in the contact gap between the ink layer and the chemical tempering layer in each region during the sintering heating process to obtain gap heat transfer difference information; based on the gap heat transfer difference information, analyzing the flow and filling characteristics of low-melting-point glass powder in the gap due to different heat transfer rates in each region to obtain melt coverage-contact state information; and based on the melt coverage-contact state information, analyzing the modulation effect of the actual ion exchange interface between the low-melting-point glass powder melt and the tempering layer on the ion exchange activation energy to obtain the secondary ion exchange information.

[0013] Optionally, the step of analyzing the changes in surface layer ion concentration in each region caused by the difference in potassium and sodium ion exchange rates, and the resulting redistribution of local compressive stress, based on the secondary ion exchange information, to obtain the defect association information set includes: analyzing the non-uniform accumulation of ion concentration during sintering and cooling process in the residual and non-residual regions due to the difference in potassium and sodium ion exchange rates, based on the secondary ion exchange information, to obtain ion concentration gradient evolution information; and analyzing the non-uniform release characteristics of local compressive stress caused by the obstruction of ion migration during the cooling stage, based on the ion concentration gradient evolution information, to obtain the defect association information set.

[0014] Optionally, generating and outputting cover plate defect discrimination information based on the defect association information set includes: analyzing the deviation information between the defect characteristics of each region and the preset cover plate quality standard based on the defect association information set to obtain the defect severity level; and generating and outputting the cover plate defect discrimination information including defect location, defect type and defect level based on the defect severity level.

[0015] Secondly, this application provides a cover plate defect discrimination system based on the fusion of cleaning and tempering screen printing data. The system includes: a watermark residue module, used to acquire an information set of irregular curved surface cover plates, and based on the information set of irregular curved surface cover plates, analyze the invisible ion watermark residue information caused by the limitation of the curved surface structure and chemical tempering to obtain a residue information set; a defect association module, used to analyze the secondary ion exchange effect between the ink low melting point glass powder and the chemical tempering layer during screen printing, and the surface stress distribution change caused by the secondary ion exchange effect, based on the residue information set to obtain a defect association information set; and a defect discrimination module, used to generate and output cover plate defect discrimination information according to the defect association information set. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application; Figure 2 A flowchart of a cover plate defect identification method based on the fusion of cleaned and tempered screen printing data provided in an embodiment of this application; Figure 3 This is a schematic diagram of a cover plate defect identification system based on the fusion of cleaning and tempering screen printing data, provided in an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0019] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0020] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0021] In the process of cleaning, tempering and screen printing of irregular curved surface covers, the production process of irregular curved surface covers is complicated. Key processes such as cleaning and tempering are prone to hidden defects. These defects only appear in the later stages. Existing manual and AOI inspections cannot identify them in advance or trace their causes. The risk of defective products flowing out is high and it is difficult to locate the process failure links.

[0022] Based on this, this application provides a method and system for identifying cover plate defects based on the fusion of cleaning, tempering, and screen printing data. By fusing data from the entire cleaning-tempering-screen printing process, the physical and chemical states of the upstream processes are quantitatively analyzed to determine the spatial mismatch between the non-uniform residue of the cleaning solution caused by the curved surface structure and the ion exchange depth of the tempering layer, thereby accurately locating potential defect sources. Furthermore, a secondary ion exchange mechanism is introduced to reveal a new round of ion migration behavior between the low-melting-point glass powder of the ink and the tempering layer at the microscopic contact interface during the screen printing sintering stage. This behavior is modulated by the previous residual state, directly leading to local stress redistribution and abnormal optical properties, thus solving the problems of difficulty in identifying hidden defects in irregular curved surface cover plates and difficulty in tracing their causes in the prior art.

[0023] Figure 1 This application provides an illustration of an application scenario. During the cleaning, tempering, and screen printing process of irregularly shaped curved cover plates, the method provided in this application is applied to shift from post-inspection to pre-inspection, significantly improving the accuracy and interpretability of defect identification.

[0024] Specifically, the method provided in this application can be applied to any server. The server interacts with the cover plate production line inspection equipment to obtain the information set of irregular curved surface cover plates provided by the cover plate production line inspection equipment, constructs a complete association between residual features and defect causes, generates and outputs cover plate defect discrimination information to cover plate production staff, significantly improves defect identification accuracy and coverage, reduces defective output, provides data basis for method iteration, and comprehensively improves cover plate processing quality and production efficiency.

[0025] The specific implementation method can be referred to in the following embodiments, wherein the data mentioned in the embodiments are only for reference and examples, so that relevant personnel can better understand them.

[0026] Figure 2This is a flowchart illustrating a cover plate defect identification method based on the fusion of cleaned tempered screen printing data, provided in one embodiment of this application. The method of this embodiment can be applied to the server in the above scenario. For example... Figure 2 As shown, the method includes: S201. Obtain the information set of the irregular curved surface cover plate. Based on the information set of the irregular curved surface cover plate, analyze the residual information of invisible ion water marks caused by the limitation of the curved surface structure and chemical tempering, and obtain the residual information set.

[0027] The information set for irregularly shaped curved cover plates can be a collection of information such as process parameters, processing environment, and material ratios generated throughout the entire process of chemical tempering, cleaning, and screen printing, with the data source being the cover plate production line testing equipment. The residual information set can be a collection of characteristic data of invisible ion water marks on the cover plate surface.

[0028] Specifically, in the field of cover plate processing, irregular curved surface structures make it difficult for cleaning fluid to flow evenly and be completely discharged in the concave and corner areas of the curved surface. Potassium and sodium ions introduced during the chemical tempering process easily combine with the cleaning water to form highly stable invisible ionic water marks. These water marks are difficult to identify by conventional detection methods, but they directly affect the forming quality of subsequent screen printing processes and the appearance performance of the cover plate. If the residual information set is not accurately extracted first, subsequent defect identification will result in missed or false identifications. Therefore, obtaining and analyzing the information set of irregular curved surface cover plates to obtain the residual information set is the basic prerequisite for achieving accurate defect identification.

[0029] S202. Based on the residual information set, analyze the secondary ion exchange effect between the low-melting-point glass powder of the ink and the chemical tempering layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect to obtain the defect correlation information set.

[0030] Low-melting-point glass powder can be a key filler in screen printing inks. The chemically tempered layer can be a high-strength surface structure formed by chemically tempering the cover plate. The secondary ion exchange effect can be the phenomenon of migration and replacement of glass powder ions with tempered layer ions during screen printing. Changes in surface stress distribution can be the state where ion exchange disrupts the uniformity of stress in the cover plate. The defect correlation information set can be the set of correspondences between watermarks, ion exchange, stress, and defects.

[0031] Specifically, in actual processing, residual invisible ion water marks can alter the ionic activity and interface state of the chemically tempered layer surface, causing the secondary ion exchange effect between the ink's low-melting-point glass powder and the chemically tempered layer during screen printing to exhibit non-uniformity. This leads to disordered surface stress distribution, and abnormal stress distribution is the direct cause of core defects such as screen printing peeling, cracks, color differences, and hidden cracks in the cover plate. Without analyzing this effect and stress changes, it is impossible to establish a correlation link between residual information and defect results, and defect identification lacks a scientific basis. Therefore, conducting the above analysis based on the residual information set is a necessary step. This analysis process uses surface stress testing technology and ion energy dispersive spectroscopy analysis, combined with the water mark characteristics in the residual information set, to qualitatively and quantitatively analyze the ion migration path and exchange degree of secondary ion exchange, while simultaneously monitoring the stress values ​​and distribution patterns at various points on the cover plate surface.

[0032] S203. Generate and output cover plate defect discrimination information based on the defect association information set.

[0033] The cover plate defect identification information can be a comprehensive result of the cover plate defect type, location, level and cause, and is used for quality judgment and process optimization.

[0034] Specifically, by using a complete defect association information set and employing pattern recognition and feature matching algorithms, the ion exchange characteristics and stress distribution characteristics actually detected on the cover plate are compared with the standard association model in the defect association information set to quickly identify the defect type and cause, accurately classify the defect level, and finally generate and output clear and intuitive defect identification information.

[0035] The above technical solution first accurately extracts residual information of invisible ion watermarks based on the information set of irregular curved cover plates, overcoming the difficulty of detecting watermarks caused by curved surfaces and chemical tempering. Then, based on the residual information, it analyzes the secondary ion exchange behavior and surface stress changes between the screen-printed glass powder and the tempered layer, constructing a complete correlation between residual features and defect causes, making up for the shortcomings of existing discrimination methods that only look at the appearance and ignore internal stress and ion effects. Finally, it outputs accurate discrimination conclusions based on defect correlation information, completing the integrated closed loop from method data to defect judgment, significantly improving the accuracy and coverage of defect identification, reducing defective output, providing data basis for method iteration, and comprehensively improving the processing quality and production efficiency of cover plates.

[0036] In some embodiments, based on the information set of the irregular curved surface cover, the residual information of invisible ion water marks caused by the curvature structure and chemical tempering is analyzed to obtain a residual information set, including: the information set of the irregular curved surface cover includes a cleaning information set, a tempering information set, and a screen printing information set; based on the cleaning information set and combined with the tempering information set, the distribution of cleaning fluid residue area caused by the change of curvature of the irregular curved surface cover during the cleaning process, and the mismatch information between the ion exchange depth and the residue area during the chemical tempering process are analyzed to obtain preliminary residual information; based on the preliminary residual information and combined with the screen printing information set, the uniformity of ink coverage and masking effect in the preliminary residue area during the screen printing process are analyzed to obtain a residual information set.

[0037] Invisible ion water mark residue information can be a set of microscopic defect information formed on the surface of an irregularly shaped curved cover plate after cleaning and chemical tempering due to process coupling. These defects are not easily observed with the naked eye but can be revealed through subsequent screen printing processes. The cleaning information set can encompass all process parameters and data of the cleaning process for the irregularly shaped curved cover plate. The tempering information set can encompass all process parameters and data of the chemical tempering process. The screen printing information set can encompass all process parameters and data of the screen printing process. The distribution of cleaning fluid residue areas can be a set of residual areas with specific locations and shapes formed on the cover plate surface after cleaning due to incomplete evaporation or drainage caused by the curvature of the surface. The mismatch information between ion exchange depth and residual areas can refer to the spatial mismatch state caused by cleaning fluid residues hindering ion migration or altering the local chemical environment during chemical tempering, resulting in a significant difference between the actual ion exchange depth of the residual areas and the non-residual areas. Preliminary residue information can be the information of spatially overlapping areas extracted after spatially comparing the distribution of cleaning fluid residue areas and the mismatch information between ion exchange depth and residual areas. Uniformity of coverage refers to the degree of uniformity in the thickness and density of the ink coating on the area indicated by the initial residual information during the screen printing process. Masking effect refers to the effect of the ink layer covering or amplifying optical defects in the underlying initial residual area after the screen printing ink has cured, due to uneven coverage or stress differences with the glass substrate.

[0038] Specifically, in the production of irregularly shaped curved cover plates, without integrating cleaning, tempering, and screen printing data, it becomes impossible to identify hidden watermarks and areas of ion exchange mismatch on the curved surface, leading to missed ink coverage defects and misjudgments of residue, directly affecting the accuracy of subsequent defect identification. To address these issues, multi-process data linkage analysis accurately extracts the characteristics of hidden residues and ink coverage defects on the curved surface, eliminating blind spots in single-stage detection and providing a complete and reliable basis for secondary ion exchange effect analysis and defect identification.

[0039] In the specific analysis process: First, a cleaning information set containing cleaning process parameters (such as cleaning fluid formulation, spray pressure, and drying temperature curve) and three-dimensional curvature data of the cover plate was acquired, along with a tempering information set containing contour maps of tempering salt bath temperature, time, and ion exchange depth. Using a pre-built fluid simulation model, based on the cleaning information set, the flow and evaporation process of the cleaning fluid in the concave (R<10mm) and convex (R>20mm) regions of the cover plate were simulated. The liquid film thickness distribution and drying time at each location were calculated, and the distribution of residual cleaning fluid areas was output in the form of a mesh map. This distribution map uses a 0.5mm×0.5mm mesh size, marking areas where the drying time exceeds 120% of the process baseline value as potential residual areas. Next, the ion exchange depth contour map from the tempering information set is retrieved and spatially registered with the aforementioned residual area distribution map. Using a spatial overlap analysis algorithm, a Boolean AND operation is performed on the grid areas where the ion exchange depth is lower than 85% of the design value and the residual area grids to extract the set of spatially overlapping grids, which serves as preliminary residual information. This information is stored in the form of a coordinate list and corresponding residual-tempering anomaly feature vectors (such as liquid film thickness and ion exchange depth difference), providing precise "targets" for the next step of analysis.

[0040] In alternative or modified implementations: For determining the residual cleaning fluid area, fluid simulation can be omitted. Instead, online optical detection (such as polarized light imaging) can be used to directly obtain the water stain morphology image of the dried cover surface and use it as the distribution of the residual cleaning fluid area. For the mismatch between ion exchange depth and residual area, in addition to using tempering process parameters for estimation, non-destructive detection methods, such as micro-area X-ray fluorescence spectroscopy (μ-XRF) scanning, can be used to directly measure the potassium ion concentration on the surface of each area. Areas with potassium ion concentrations below the average value are defined as mismatched areas. When generating preliminary residual information, the spatial registration algorithm can also be upgraded to a probability-based Bayesian fusion model, comprehensively considering image alignment errors and measurement noise, and outputting preliminary residual information with confidence.

[0041] In some embodiments, the process of constructing preliminary residual information includes: based on the cleaning information set, analyzing the liquid film thickness distribution formed in the concave and convex regions of the irregular curved cover plate due to the difference in surface tension of the cleaning fluid and the different drying rates, to determine the cleaning fluid residual information; based on the tempering information set, analyzing the difference in diffusion rate of ion exchange depth in the planar region and the curvature change region during chemical tempering due to the stress gradient, to obtain ion exchange distribution information; comparing the cleaning fluid residual information with the ion exchange distribution information, and extracting the spatially overlapping region where the ion exchange depth is lower than the depth of the residual region as the preliminary residual information.

[0042] The surface tension difference of the cleaning solution can be attributed to the difference in surface tension between the concave and convex areas of an irregularly shaped curved cover plate, caused by the different contact angles at the interfaces. The drying rate refers to the speed at which the cleaning solution completely evaporates from a liquid to a gaseous state in different curvature areas of the cover plate. The chemical tempering process can be a glass strengthening process that uses high-temperature molten salt to exchange potassium and sodium ions on the glass surface, forming a compressive stress layer. The liquid film thickness distribution can be the spatial distribution of the liquid film thickness on the concave and convex surfaces of the cover plate after cleaning, due to differences in surface tension and drying rate. The cleaning solution residue information can be the location and extent of areas that are not completely dried after cleaning and still have a liquid film covering them. The ion exchange distribution information can be the distribution of exchange depth caused by the different ion diffusion rates in planar and curved areas during tempering due to stress gradients. The stress gradient can be the amplitude and spatial trend of stress changes from the glass surface to the interior during chemical tempering. The ion exchange depth can be the vertical penetration thickness at which potassium and sodium ions effectively exchange within the glass.

[0043] Specifically, during the initial residual information construction process, without analyzing surface tension differences, drying rates, and stress gradients, it's impossible to identify the overlapping areas of liquid film residue and ion exchange mismatch, leading to missed detection of hidden watermark areas and subsequent screen printing defects. To address this issue, by fully coupling the differences in key parameters between cleaning and tempering, the risk area of ​​hidden residues is accurately located, providing a reliable benchmark for subsequent screen printing masking effects and secondary ion exchange analysis, significantly reducing misjudgments and missed detections of defects in irregularly shaped cover plates.

[0044] In the specific analysis process: First, finite element fluid dynamics simulation was used to simulate the flow and drying process of the cleaning fluid in the concave and convex areas of the CAD model of the irregularly shaped curved cover plate. The boundary conditions of the simulation model were provided by the cleaning information set, including the cleaning fluid viscosity, surface tension coefficient (based on the cleaning fluid formula), oven temperature, and wind speed. The simulation output was the distribution of liquid film thickness at various points on the cover plate surface. Then, by correlating the liquid film thickness with the cleaning fluid concentration, the residual information of the cleaning fluid was quantified. Simultaneously, through stress-diffusion coupled finite element analysis, based on the tempering information set (including molten salt temperature, time, and composition), and importing the stress distribution data of the cover plate during the chemical tempering process, a diffusion model was established. In this model, the stress gradient was reflected by modifying the activation energy term of the diffusion equation—specifically, a higher diffusion activation energy was set in areas with large stress gradients (such as outside the R-angle). The simulation output of this model was the spatial distribution of the ion exchange depth, i.e., the ion exchange distribution information. Finally, in a unified three-dimensional mesh coordinate system of the cover plate, the depth field of the residual cleaning fluid information was compared node by node with the depth field of the ion exchange depth. By traversing all grid nodes, the set of all nodes with "cleaning fluid residual depth value > ion exchange depth value" is extracted. The continuous spatial region formed by these nodes is the high-risk area under the combined action of the intrinsically safe "wet method + thermal method", and is defined as the preliminary residual information.

[0045] In alternative or modified implementations: if a precise physical simulation model is lacking, an experimental calibration + interpolation method can be used. This involves selecting test pieces with typical curvature characteristics (such as concave and convex test pieces with different radius angles), measuring the residual cleaning fluid amount using precise weighing or surface composition analysis (such as XPS), and measuring the ion exchange depth using electron probe microanalysis (EPMA), thus constructing an empirical mapping table between curvature and residual amount / depth. In practical applications, for each region of the irregularly shaped curved cover plate, based on its local curvature, corresponding cleaning fluid residual information and ion exchange distribution information are generated by looking up tables and interpolation. Furthermore, for the step of comparing and extracting spatially overlapping regions, if the data resolutions of the two types of information are different, the ion exchange distribution information can be interpolated to the same grid resolution as the cleaning fluid residual information using least-squares fitting or Kriging interpolation before comparison, and then a point-by-point comparison can be performed.

[0046] In some embodiments, the process of constructing ion exchange distribution information includes: based on the tempering information set, analyzing the stress distribution differences caused by the different curvature of the irregular curved surface cover plate in the planar region and the curvature variation region, to obtain tempering stress distribution information; based on the tempering stress distribution information, analyzing the modulation direction and modulation amplitude of the stress gradient in each region on the ion diffusion activation energy, to obtain ion exchange distribution information.

[0047] The stress distribution information can be the data on the magnitude and distribution of stress in the planar and curvature variation areas of an irregularly shaped curved cover plate after chemical tempering, due to the difference in surface curvature. The ion diffusion activation energy can be the minimum energy required for ions to complete diffusion and migration within the glass layer. The modulation direction and amplitude can be the direction in which the stress gradient increases or decreases the activation energy, and the numerical range within which the activation energy is changed.

[0048] Specifically, in the process of constructing ion exchange distribution information, failure to analyze the stress differences caused by curvature can lead to distorted activation energy calculations, incorrect ion exchange rate judgments, and missed defects in curved areas, directly reducing the accuracy of subsequent defect identification. To address these issues, activation energy calculations are precisely modulated by stress gradients, reliably restoring the true ion exchange state at various points on the curved surface. This provides an accurate basis for residual area matching and defect correlation analysis, significantly improving the stability of defect identification in irregularly shaped cover plates.

[0049] In the specific analysis process: First, the tempering information set is invoked. This dataset contains the 3D geometric model of the irregular curved cover plate (including surface curvature radius data), chemical tempering process parameters (such as potassium nitrate molten salt temperature, tempering time, and molten salt concentration), and material properties (such as the elastic modulus and Poisson's ratio of glass). The system uses finite element analysis software (such as ANSYS or Abaqus) to establish a thermo-solid coupling model of the irregular curved cover plate. By loading tempering process parameters (such as 400℃ constant temperature treatment for 4 hours), the system simulates and calculates the differences in thermal stress and structural stress caused by different surface curvatures (e.g., a convex surface with R=50mm and a flat surface with R=100mm), and outputs tempering stress distribution information containing stress values ​​at each node. Subsequently, based on this stress distribution information, the stress gradient tensor (i.e., the rate of change of stress in the x, y, and z directions) at each finite element node is extracted. Based on a pre-calibrated stress-activation energy modulation model (this model establishes the relationship between the stress gradient magnitude and K through molecular dynamics simulation or experimental determination), the system calculates the stress gradient magnitude and K. + / Na + The functional relationship between the changes in activation energy for ion diffusion, such as ΔEa=k*| σ|, where k is the modulation coefficient, σ represents the stress gradient), and the actual activation energy of ion diffusion in each region is calculated (Ea_actual = Ea_base + ΔEa). Finally, based on Fick's second law, this actual activation energy value drives the ion diffusion simulation, obtaining the ion concentration distribution curve at each location along the tempering depth direction, thereby constructing ion exchange distribution information, which is a data volume containing "(x,y,z) coordinates - ion exchange depth - potassium ion concentration".

[0050] In alternative or modified implementations, the means of obtaining tempering stress distribution information are not limited to finite element simulation. For example, a photoelastic stress tester can be used to scan the tempered cover plate sample to directly obtain the birefringence fringe distribution in the plane and curvature variation regions. The actual residual stress at each point can be inverted using photoelastic theory, which serves as the source of tempering stress distribution information. In addition, when establishing a stress-activation energy modulation model, besides using the above-mentioned functional relationship, a black-box prediction model can also be trained based on a large amount of experimental data using machine learning algorithms (such as support vector regression or Gaussian process regression) to directly predict the corresponding ion exchange depth according to the stress gradient, thereby skipping the intermediate step of activation energy calculation and directly obtaining ion exchange distribution information.

[0051] In some embodiments, based on preliminary residual information and combined with a screen printing information set, the uniformity of ink coverage and masking effect in the preliminary residual area during screen printing are analyzed to obtain a residual information set, including: based on the preliminary residual information, analyzing the local micro-separation state caused by the difference in shrinkage stress during the curing and shrinkage of the ink to obtain coverage uniformity information; based on the coverage uniformity information and combined with the screen printing information set, analyzing the influence of the stress difference between the residual area and the non-residual area of ​​the cured ink layer on the optical transmission characteristics to obtain masking effect information; and integrating the coverage uniformity information and the masking effect information to construct a residual information set.

[0052] The screen printing process can be described as the process of printing ink onto the surface of an irregularly shaped curved cover plate using a screen printing plate. The curing and shrinkage process can be described as the process by which ink changes from a liquid to a solid state and undergoes volume shrinkage under heating and ultraviolet light. Shrinkage stress differences can be described as the uneven stress distribution caused by the different shrinkage rates between the residual and non-residual areas during ink curing. The ink layer can be described as the thin film layer formed by the cured ink covering the cover plate surface after screen printing. Optical transmission characteristics can be described as the optical transmission performance of the cover plate, including transmission, refraction, and scattering of visible light. Coverage uniformity information can be described as quantitative characteristic information regarding the uniformity of ink coverage in the residual areas. Masking effect information can be described as characteristic information regarding the obstruction and interference of optical transmission caused by stress differences in the ink layer. Local micro-separation state can be described as the physical state in which tiny gaps appear between the ink and the tempered layer due to uneven shrinkage.

[0053] Specifically, during the screen printing and curing shrinkage process, failure to analyze the differences in shrinkage stress and the masking effect can lead to the omission of local micro-separation defects, resulting in distorted residual information and directly causing subsequent defect misjudgments and a decrease in cover plate yield. To address these issues, by accurately extracting the characteristics of coverage uniformity and masking effect, the true state of the residual area is fully restored, ensuring the accuracy and reliability of the residual information set and laying a solid foundation for subsequent secondary ion exchange analysis and defect identification.

[0054] In the specific analysis process: First, the ink characteristic parameters (such as curing shrinkage rate and elastic modulus) from the preliminary residual information and screen printing information are received. Based on the ink characteristic parameters and the geometric model of the cover plate surface, a thermal stress simulation model is established. For each residual area marked by the preliminary residual information, the simulation model simulates the distribution of shrinkage stress field caused by the difference in thermal expansion coefficients between the residue below (such as cleaning liquid or tempered salt stains) and the glass substrate during the ink curing and cooling process. When the calculated shrinkage stress at a certain local point exceeds the theoretical adhesion strength threshold between the ink layer and the glass substrate, it is determined that local micro-separation has occurred at that point, and the spatial coordinates and quantitative values ​​of the separation degree (such as gap height) of all micro-separation points are recorded. These data are summarized into coverage uniformity information. Subsequently, based on this coverage uniformity information and combined with the intrinsic optical parameters of the ink layer, the influence of each micro-separation area on the transmittance, reflectance, and scattering angle distribution of incident light is analyzed using a ray tracing algorithm. The algorithm quantifies the attenuation of visual features (such as contrast and edge sharpness) in the underlying residual region by comparing the optical response differences between regions with and without micro-separation (i.e., regions that are perfectly aligned). This generates a masking coefficient distribution map, i.e., masking effect information. Finally, the spatially resolution-consistent coverage uniformity information and masking effect information are fused at the data level, for example, through data overlay or feature concatenation, to form a multidimensional tensor, which serves as the final residual information set output.

[0055] In alternative or modified implementations: the thermal stress simulation model can be replaced by analytical formulas or machine learning models based on measured data to speed up calculations. For example, for a specific batch of ink, an empirical formula for the ink curing shrinkage rate and measured micro-separation density can be established through pre-experiments. For the extraction of masking effect information, in addition to ray tracing, a modified model of the Cauchy dispersion formula or the Beer-Lambert law can also be used to calculate optical properties based on the changes in ink layer thickness and refractive index caused by stress. In the data fusion stage, if the data resolutions of coverage uniformity information and masking effect information are different, bilinear interpolation or downsampling can be performed on either information first to align their resolutions before fusion. In addition, the fusion method is not limited to data superposition. A lightweight neural network can also be constructed to take the two pieces of information as input features and output a comprehensive visibility-weighted residual information.

[0056] In some embodiments, based on the residual information set, the secondary ion exchange effect between the low-melting-point glass powder of the ink and the chemically tempered layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect are analyzed to obtain a defect-related information set. This includes: based on the residual information set, analyzing the interfacial contact gap caused by the local micro-separation state of the low-melting-point glass powder of the ink in the residual area, and the contact characteristics formed by stress differences in the non-residual area, to obtain interfacial contact characteristic information; based on the interfacial contact characteristic information, analyzing the differences in ion exchange activation energy caused by different contact states between the ink layer and the chemically tempered layer in each region during sintering, to obtain secondary ion exchange information; based on the secondary ion exchange information, analyzing the changes in surface layer ion concentration in each region caused by differences in potassium and sodium ion exchange, and the resulting redistribution of local compressive stress, to obtain a defect-related information set.

[0057] A chemically tempered layer can be an ion-exchange reinforced layer with high compressive stress resistance formed on the surface of glass after chemical tempering treatment. Changes in surface stress distribution can be caused by alterations in the magnitude, gradient, and uniformity of surface compressive stress in the chemically tempered layer due to secondary ion exchange. The sintering process can be the heat treatment process of ink melting, filling, and bonding with the glass interface during the high-temperature curing stage after screen printing. Differences in ion exchange activation energy can be due to the different energy barriers that potassium and sodium ions need to overcome to exchange under different interfacial contact states. Differences in potassium and sodium ion exchange can be due to differences in the diffusion rate, exchange depth, and concentration of potassium and sodium ions between residual and non-residual regions. Local compressive stress redistribution can be caused by the non-uniform accumulation of ion concentration, leading to a redistribution of local compressive stress in the tempered layer and stress distortion.

[0058] Specifically, in analyzing the secondary ion exchange and stress changes between the ink and the tempered layer, if the differences in interface contact and activation energy are not quantified, uneven potassium-sodium exchange cannot be identified, directly leading to localized stress distortion, missed detection of hidden defects, and ultimately, cover plate breakage. To address these issues, by accurately establishing the correlation link between interface contact, ion exchange, and stress redistribution, the causes of defects can be accurately located, significantly improving the accuracy of hidden defect identification and preventing cover plate defects and failures caused by stress distortion.

[0059] In the specific analysis process: First, the residual information set constructed through the aforementioned steps is read. This information set accurately marks the residual and non-residual areas to which each pixel on the cover plate surface belongs. Based on this, a microscopic contact state analysis process is initiated. For the residual areas, local micro-separation state characteristic parameters are extracted. These parameters characterize the distribution, width (e.g., 0.1-0.5 micrometer gaps calibrated by electron microscopy), and spatial density of the tiny gaps between the ink layer and the tempered layer caused by the shrinkage stress of ink curing. For the non-residual areas, the system extracts the contact features formed by stress differences. That is, due to the stress differences on the tempered layer surface and the more uniform shrinkage of the ink layer, this is manifested as tight contact or gapless contact. Subsequently, these features are encapsulated into interface contact feature information.

[0060] Next, based on the interface contact characteristic information, a sintering process ion exchange model is invoked. The core of this model is a thermo-chemical coupled finite element solver. For the gap portion of the residual region, the solver first calculates the thermal expansion behavior of the residual gas in the gap during the sintering heating process (e.g., from room temperature to 650°C) to obtain the gap heat transfer difference information. That is, the existence of the gap significantly reduces the heat conduction efficiency, resulting in the lag of ink melting and interface heating in this region. Based on this lag effect, the solver calculates the flow and filling characteristics of low melting point glass powder in the gap: due to the temperature lag, the melt viscosity is high, making it difficult to completely fill the tiny gaps, forming local uncontacted areas. This is the melt coverage-contact state information. Finally, the solver uses the Arrhenius equation, taking the melt coverage-contact state information of each region (i.e., the actual ion exchange contact area and local temperature) as input, to calculate the modulation amplitude of the ion exchange activation energy caused by the contact state difference in this region, and outputs secondary ion exchange information.

[0061] In alternative or modified implementations: The micro-contact state analysis process of this embodiment can adopt online detection technology based on light scattering or surface plasmon resonance to replace the direct extraction based on microstructure parameters, so as to monitor the changes in interface gaps in real time during sintering in a non-contact manner. In addition, the finite element solver in the ion exchange model of the sintering process can be replaced with a model based on molecular dynamics simulation. This model can more accurately describe the ion exchange dynamics of the interface between the low melting point glass powder melt and the tempered layer at the atomic scale, thereby providing more accurate activation energy modulation prediction, especially when dealing with new ink formulations or special glass substrates.

[0062] In some embodiments, the process of constructing secondary ion exchange information includes: based on interface contact characteristic information, analyzing the thermal expansion behavior of residual gas in the contact gap between the ink layer and the chemical tempering layer in each region during the sintering heating process to obtain gap heat transfer difference information; based on the gap heat transfer difference information, analyzing the flow and filling characteristics of low-melting-point glass powder in the gap due to different heat transfer rates in each region to obtain melt coverage-contact state information; and based on the melt coverage-contact state information, analyzing the modulation effect of the actual ion exchange interface between the low-melting-point glass powder melt and the tempering layer on the ion exchange activation energy to obtain secondary ion exchange information.

[0063] Interface contact characteristics can include contact attributes such as the contact gap and tightness between the ink layer and the chemical tempering layer in different areas. Residual gas within the contact gap can be air, water vapor, or other gaseous substances remaining in the contact gap between the ink and the tempering layer. The sintering heating process can be the process of heating and curing the cover plate after screen printing, causing the low-melting-point glass powder to melt and rise in temperature. Thermal expansion behavior can be the physical changes of residual gas within the gap, such as volume expansion and pressure increase, as the sintering temperature rises. Gap heat transfer difference information can be the difference in heat transfer rate caused by the different degrees of gas expansion in different gaps. Flow filling characteristics can be the distribution characteristics of the low-melting-point glass powder melt within the gap, such as the flow path and filling fullness. Melt coverage-contact state information can be the coverage and adhesion state of the melt with the tempering layer after filling the gap. The actual ion exchange interface of the tempering layer can be the actual contact interface where ion exchange occurs between the low-melting-point glass powder melt and the tempering layer. Ion exchange activation energy can be the minimum energy required for potassium and sodium ions to undergo an exchange reaction at the interface. Modulation effect can be the change and influence of the actual ion exchange interface on the ion exchange activation energy.

[0064] Specifically, in the analysis of secondary ion exchange effects, failure to quantify the differences in gas thermal expansion and heat transfer can lead to distorted judgments of melt filling and deviations in activation energy calculations, directly causing errors in ion exchange analysis and resulting in missed or misjudged defects in the cover plate. To address these issues, by accurately reproducing the true state of interstitial heat transfer and melt filling during sintering, and precisely quantifying the activation energy modulation effect, the accuracy of secondary ion exchange information is significantly improved, providing reliable data support for subsequent defect correlation analysis.

[0065] In the specific analysis process: the obtained interface contact feature information is used as input. For example, for an irregular curved cover plate, the interface contact feature information shows that the ink-tempered layer contact gap width in the concave area is 1-3 micrometers, and nitrogen and a small amount of water vapor are sealed in the gap, while the contact gap in the planar area is 0.1-0.5 micrometers, and the gas is rarefied. The secondary ion exchange information construction process for this cover plate is initiated. First, the heating rate corresponding to this cover plate product (such as heating from room temperature to 680℃ at 15℃ / min) is obtained through the pre-established sintering heating curve database. Then, based on the ideal gas law and microfluidic thermodynamic model, the thermal expansion behavior of the sealed gas in the narrow gap in the concave area during the heating process is analyzed. It is calculated that when the temperature rises to 300℃, the gas pressure in the gap rapidly increases to 3 times the ambient pressure. This high pressure hinders the flow of external low-melting-point glass powder melt into the depth of the gap. In the planar region, due to the extremely small gaps and thin gas, the thermal expansion effect of the gas is negligible, allowing the melt to rapidly spread and wet the tempered layer surface, thus generating information on gap heat transfer differences. In the concave region, due to the presence of a high-pressure gas layer, the thermal conductivity is only 20% of that in the planar region. Next, using computational fluid dynamics (CFD), the characteristics of glass powder melting and flow under different heat transfer conditions are simulated. The simulation results show that in the planar region, the melt completely covers the tempered layer at 600℃; while in the concave region, the melt only fills the gap inlet, forming a melt-coverage-contact state. That is, the concave region is in a partially non-contact state, while the planar region is in a fully contact state. Finally, based on these two distinct contact states, the ion exchange activation energy modulation function library is invoked to map the fully contact state to an exchange parameter with a 30% reduction in activation energy, and the partially non-contact state to an exchange parameter with a 15% increase in activation energy, thereby outputting accurate secondary ion exchange information.

[0066] In alternative or modified implementations: the gas composition relied upon for analyzing the thermal expansion behavior of gas within the gap can be provided not only by interfacial contact characteristic information but also inferred by combining curing process parameters of the screen printing ink (such as curing temperature and time). For example, if it is known that the ink produces trace amounts of organic gas when cured at 120°C, this gas component needs to be included in the calculation during modeling. Furthermore, for the simulation of the melt-coverage-contact state, if computational resources are limited, a complete CFD simulation can be abandoned, and instead, empirical formulas based on capillary number and Weber number can be used to quickly assess the melt filling depth and spreading area, thereby constructing approximate melt-coverage-contact state information. When analyzing the modulation effect of ion exchange activation energy, in addition to contact area and tightness, the surface energy parameter between the glass powder melt and the tempered layer can be introduced. This parameter is used to correct for the activation energy change caused by differences in melt wettability.

[0067] In some embodiments, based on secondary ion exchange information, the changes in surface layer ion concentration in each region caused by differences in potassium and sodium ion exchange, and the resulting redistribution of local compressive stress, are analyzed to obtain a defect association information set. This includes: based on secondary ion exchange information, analyzing the non-uniform accumulation of ion concentration in the residual and non-residual regions during the sintering cooling process due to differences in potassium and sodium ion exchange rates, to obtain ion concentration gradient evolution information; and based on ion concentration gradient evolution information, analyzing the non-uniform release characteristics of local compressive stress caused by hindered ion migration during the cooling stage, to obtain a defect association information set.

[0068] Information on the evolution of ion concentration gradients can be information on the spatial distribution changes of ion concentration caused by the difference in potassium and sodium ion exchange rates during the cooling stage. The sintering cooling process can be the process of the cover plate gradually cooling from a high temperature to room temperature after the ink sintering is completed. Non-uniform accumulation can be the uneven aggregation phenomenon of ion concentration between the residual and non-residual areas due to the difference in exchange rates. Ion migration obstruction can be the state in which the glass viscosity increases during the cooling stage, resulting in the restriction of the diffusion and movement of potassium and sodium ions. The non-uniform release characteristics of local compressive stress can be the differentiated release of surface compressive stress in different regions caused by the obstruction of ion migration.

[0069] Specifically, during the sintering and cooling process, failure to analyze the non-uniform accumulation and hindered migration of ions, as well as stress release, will directly lead to the omission of hidden stress defects, resulting in insufficient strength and breakage of the cover plate, severely impacting product yield and safety. To address these issues, by accurately capturing abnormal ion concentration and stress distribution during the cooling stage, hidden stress defects can be fully identified, significantly improving the reliability of the identification and providing an accurate basis for subsequent quality grading and method optimization.

[0070] In the specific analysis process: First, secondary ion exchange information is obtained, which includes the difference in ion exchange activation energy of each microscopic spatial element (e.g., a 5 μm × 5 μm grid cell) and the corresponding interface contact state. This data is then input into a physical mechanism-based diffusion-reaction dynamic simulation engine. Instead of using simple formulas, this engine simulates ion migration behavior under cooling temperature curves through numerical iteration. The engine first calculates the weighting coefficients of the diffusion capacity of potassium and sodium ions at different temperatures based on the activation energy difference of each grid cell. Then, driven by the cooling temperature curve (e.g., a time-temperature drop sequence from 720°C to room temperature), the engine progressively calculates the net diffusion of potassium and sodium ions between grid cells and along the thickness direction within each tiny time step. During the migration process, the engine tracks the potassium to sodium ion concentration ratio at each location in real time and records its change over time, forming an ion concentration ratio evolution map covering the entire cover glass surface. This is the ion concentration gradient evolution information. Next, this evolution information is spatially correlated with the physical properties of the cover glass (such as thermal expansion coefficient, elastic modulus, and viscosity-temperature relationship). By comparing the concentration ratio differences between adjacent grid cells at the same time point, it can be determined which areas have significantly faster or slower ion concentration changes than their surrounding areas. In particular, grid cells whose ion concentration ratio difference exceeds a preset threshold at the end of cooling (e.g., when the temperature drops below 400°C) are marked, indicating that ion migration is hindered at these locations during cooling. Finally, all marked abnormal cells and their corresponding stress release rate deviations (calculated from the concentration gradient intensity using a built-in empirical mapping table) are integrated to form a final dataset containing defect locations and stress anomaly levels—the defect correlation information set.

[0071] In alternative or modified implementations: the aforementioned diffusion-reaction dynamic simulation engine can be constructed using the phase-field method numerical framework, treating the ion exchange process as an interface migration between different phases (such as potassium-rich and sodium-rich phases), which is more suitable for handling complex discontinuous diffusion behavior at the boundary between residual and non-residual regions. Furthermore, to accelerate computation and meet the real-time requirements of high-speed production lines, a large number of ion concentration gradient evolution samples under different process parameter combinations can be generated offline using multiphysics simulation software in advance, and a lightweight deep neural network (DNN) can be trained as a surrogate model. In actual production, only the real-time secondary ion exchange information and cooling temperature curve need to be passed as input to the DNN, and the ion concentration gradient evolution information and the corresponding local compressive stress non-uniform release feature mapping can be directly output within milliseconds, eliminating the time-consuming iterative calculation process. Another modification is that random perturbation factors can be introduced during the simulation process to simulate microscopic impurities or temperature fluctuations that are difficult to avoid in actual processes, performing Monte Carlo simulations on the concentration gradient and stress release, thereby outputting a defect correlation information set with confidence intervals.

[0072] In some embodiments, based on the defect association information set, cover plate defect discrimination information is generated and output, including: based on the defect association information set, analyzing the deviation information between the defect characteristics of each area and the preset cover plate quality standard to obtain the defect severity level; and based on the defect severity level, generating and outputting cover plate defect discrimination information including defect location, defect type and defect level.

[0073] The defect location can be a specific spatial area on the cover plate where there is abnormal ion concentration or uneven stress. The defect type can be a variety of defects such as invisible watermarks or uneven stress caused by secondary ion exchange and stress redistribution. The defect level can be a specific quantitative indicator of the severity of the defect, used to distinguish the degree of harm caused by the defect.

[0074] Specifically, in the defect identification and output process, if the quality standard is not used to determine the level and output the location, type, and level information, it will be impossible to locate hidden defects or distinguish the degree of harm, directly leading to the wrong release of products and the inability to accurately optimize the process. To address the above problems, by accurately outputting the location, type, and level of defects, fully covering hidden defects, a clear basis is provided for product sorting and process correction, significantly reducing the rate of false positives and false negatives, and improving the accuracy of cover plate quality control.

[0075] In the specific analysis process: First, a pre-stored cover plate quality standard database is loaded. This database is derived from the client's tolerance limits for various defects and is stored in the form of a configuration file. Taking a 3D curved cover plate for a mobile phone as an example, its quality standard stipulates that: in the planar area, the transmittance reduction caused by ion water marks should not exceed 5%, and there should be no visible stripes; in the R-corner area with a large curvature, the transmittance deviation caused by uneven stress should not exceed 8%, and the ink adhesion grade should not be lower than 5B. Subsequently, the constructed defect association information set is called. This information set contains the ion concentration gradient evolution information and local compressive stress redistribution data obtained from the aforementioned analysis. The defect characteristics of each area in the defect association information set (such as the gridded planar area, R-corner area, and side area) - such as local compressive stress value, color offset, film thickness deviation - are compared one by one with the threshold of the corresponding area in the quality standard to calculate the deviation. For example, if the transmittance deviation of the R-corner area is detected to be 6% (the quality requirement is 8%), and the adhesion grade is 4B (the quality requirement is 5B), it is determined that both parameters do not exceed the standard, but one of them deviates from the standard value. Based on preset fuzzy logic rules (e.g., if one parameter is close to the threshold, it is judged as a minor defect; if both parameters are close to the threshold, it is judged as a moderate defect; if any parameter exceeds the threshold, it is judged as a severe defect), the severity level of the defect in the R-corner region is calculated to be moderate. Finally, a structured data packet is encapsulated, containing the precise spatial coordinates of the defect (e.g., the R-corner region defined based on XYZ axis coordinates), the determined defect type (e.g., uneven stress - decreased adhesion), and the corresponding defect level (e.g., moderate), as the complete cover plate defect discrimination information output.

[0076] In alternative or modified implementations, the classification of defect severity levels is not limited to simple rule-based methods. For example, machine learning models such as Support Vector Machines (SVM) or Random Forests can be used. Multidimensional feature vectors (such as stress gradients, ion concentration change rates, and optical color difference values) from the defect association information set can be used as input, along with preset cover plate quality standards as labels. Supervised learning can then be performed to train a discrimination model that can automatically output continuous probability values. The probability values ​​output by this model can be directly mapped to more refined severity levels (e.g., 0-1 points, with higher scores indicating greater severity), rather than just discrete categories of slight, moderate, and severe. Furthermore, the cover plate quality standard database itself can be dynamically updated. For instance, if the first-piece inspection of a production batch reveals that existing standards are too stringent, leading to low yield, engineers can adjust the threshold in the configuration file, increasing the upper limit of transmittance deviation tolerance in the R-corner area from 8% to 10%, thereby dynamically adjusting the discrimination standard to meet yield optimization needs.

[0077] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0078] Figure 3 This is a schematic diagram of a cover plate defect discrimination system based on the fusion of cleaning and tempering screen printing data provided in an embodiment of this application, as shown below. Figure 3 As shown, the cover plate defect discrimination system 300 based on the fusion of cleaning and tempered screen printing data in this embodiment includes: a watermark residue module 301, a defect association module 302, and a defect discrimination module 303.

[0079] The watermark residue module 301 is used to acquire the information set of the irregular curved surface cover plate, and based on the information set of the irregular curved surface cover plate, analyze the invisible ion watermark residue information caused by the limitation of the curved surface structure and chemical tempering to obtain the residue information set; the defect association module 302 is used to analyze the secondary ion exchange effect between the low melting point glass powder of the ink and the chemical tempering layer during screen printing, and the surface stress distribution change caused by the secondary ion exchange effect, based on the residue information set to obtain the defect association information set; the defect discrimination module 303 is used to generate and output the cover plate defect discrimination information according to the defect association information set.

[0080] Optionally, when the watermark residue module 301 analyzes the invisible ion watermark residue information caused by the curvature structure limitation and chemical tempering based on the irregular curved surface cover plate information set to obtain the residue information set, it is specifically used for: the irregular curved surface cover plate information set includes a cleaning information set, a tempering information set, and a screen printing information set; based on the cleaning information set and combined with the tempering information set, analyzing the distribution of cleaning fluid residue area caused by the curvature change of the irregular curved surface cover plate during the cleaning process, and the mismatch information between the ion exchange depth and the residue area during the chemical tempering process, to obtain preliminary residue information; based on the preliminary residue information and combined with the screen printing information set, analyzing the coverage uniformity and masking effect of ink in the preliminary residue area during the screen printing process, to obtain the residue information set.

[0081] Optionally, the watermark residue module 301, during the construction of the preliminary residue information, is specifically used for: analyzing the liquid film thickness distribution formed by the difference in surface tension of the cleaning fluid and the different drying rates in the concave and convex regions of the irregular curved cover plate based on the cleaning information set, and determining the cleaning fluid residue information; analyzing the difference in diffusion rate of ion exchange depth in the planar region and the curvature change region during chemical tempering due to stress gradient based on the tempering information set, and obtaining ion exchange distribution information; comparing the cleaning fluid residue information with the ion exchange distribution information, and extracting the spatially overlapping region where the ion exchange depth is lower than the depth of the residue region, as the preliminary residue information.

[0082] Optionally, the watermark residue module 301, during the construction of the ion exchange distribution information, is specifically used to: analyze the stress distribution differences caused by the different curvatures of the irregular curved surface cover plate in the planar region and the curvature change region based on the tempering information set, and obtain tempering stress distribution information; and analyze the modulation direction and modulation amplitude of the stress gradient in each region on the ion diffusion activation energy based on the tempering stress distribution information, and obtain the ion exchange distribution information.

[0083] Optionally, when the watermark residue module 301 analyzes the coverage uniformity and masking effect of ink in the initial residue area during screen printing based on the preliminary residue information and the screen printing information set to obtain the residue information set, it is specifically used for: analyzing the local micro-separation state caused by the difference in shrinkage stress during the curing and shrinking process of the ink based on the preliminary residue information to obtain coverage uniformity information; analyzing the influence of the stress difference between the residue area and the non-residue area of ​​the cured ink layer on the optical transmission characteristics based on the coverage uniformity information and the screen printing information set to obtain masking effect information; and integrating the coverage uniformity information and the masking effect information to construct the residue information set.

[0084] Optionally, when the defect association module 302 analyzes the secondary ion exchange effect between the ink low-melting-point glass powder and the chemical tempering layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect, based on the residual information set, to obtain the defect association information set, it is specifically used to: analyze the interface contact gap caused by the local micro-separation state of the ink low-melting-point glass powder in the residual area, and the contact characteristics formed by stress differences in the non-residual area, based on the residual information set, to obtain interface contact characteristic information; analyze the differences in ion exchange activation energy caused by different contact states between the ink layer and the chemical tempering layer in each region during sintering, based on the interface contact characteristic information, to obtain secondary ion exchange information; and analyze the changes in surface layer ion concentration in each region caused by differences in potassium and sodium ion exchange, and the resulting redistribution of local compressive stress, based on the secondary ion exchange information, to obtain the defect association information set.

[0085] Optionally, the defect association module 302, during the construction of the secondary ion exchange information, is specifically used to: analyze the thermal expansion behavior of residual gas in the contact gap between the ink layer and the chemical tempering layer in each region during the sintering heating process based on the interface contact feature information, and obtain gap heat transfer difference information; analyze the flow and filling characteristics of low-melting-point glass powder in the gap due to different heat transfer rates in each region based on the gap heat transfer difference information, and obtain melt coverage-contact state information; and analyze the modulation effect of the actual ion exchange interface between the low-melting-point glass powder melt and the tempering layer on the ion exchange activation energy based on the melt coverage-contact state information, and obtain the secondary ion exchange information.

[0086] Optionally, when the defect association module 302 analyzes the changes in surface layer ion concentration in each region caused by differences in potassium and sodium ion exchange rates, and the resulting redistribution of local compressive stress, based on the secondary ion exchange information, to obtain the defect association information set, it is specifically used to: analyze the non-uniform accumulation of ion concentration in the residual and non-residual regions during the sintering cooling process due to differences in potassium and sodium ion exchange rates, based on the secondary ion exchange information, to obtain ion concentration gradient evolution information; and analyze the non-uniform release characteristics of local compressive stress caused by hindered ion migration during the cooling stage, based on the ion concentration gradient evolution information, to obtain the defect association information set.

[0087] Optionally, when the defect discrimination module 303 generates and outputs cover plate defect discrimination information based on the defect association information set, it is specifically used to: analyze the deviation information between the defect characteristics of each area and the preset cover plate quality standard based on the defect association information set to obtain the defect severity level; and generate and output the cover plate defect discrimination information including defect location, defect type and defect level based on the defect severity level.

[0088] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A method for identifying defects in cover plates based on the fusion of cleaned and tempered screen printing data, characterized in that, include: Obtain the information set of the irregular curved surface cover plate, and based on the information set of the irregular curved surface cover plate, analyze the residual information of invisible ion water marks caused by the limitation of the curved surface structure and chemical tempering to obtain the residual information set; Based on the residual information set, the secondary ion exchange effect between the low-melting-point glass powder of the ink and the chemical tempering layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect, are analyzed to obtain a defect correlation information set. Based on the defect association information set, cover plate defect discrimination information is generated and output.

2. The method according to claim 1, characterized in that, Based on the information set of the irregular curved surface cover plate, the residual information of invisible ion water marks caused by the limitations of the curved surface structure and chemical tempering is analyzed to obtain the residual information set, including: The information set for the irregular curved surface cover plate includes a cleaning information set, a tempering information set, and a screen printing information set. Based on the cleaning information set and the tempering information set, the distribution of cleaning fluid residue area caused by the change of curvature of the surface during the cleaning process of the irregular curved cover plate, as well as the mismatch between the ion exchange depth and the residue area during the chemical tempering process, are analyzed to obtain preliminary residue information. Based on the preliminary residue information and the screen printing information set, the uniformity of ink coverage and masking effect in the preliminary residue area during the screen printing process are analyzed to obtain the residue information set.

3. The method according to claim 2, characterized in that, The process of constructing the preliminary residual information includes: Based on the cleaning information set, the distribution of liquid film thickness formed in the concave and convex regions of the irregular curved cover plate due to the difference in surface tension of the cleaning fluid and the different drying rates is analyzed to determine the residual information of the cleaning fluid. Based on the tempering information set, the difference in diffusion rate caused by stress gradient in the ion exchange depth during chemical tempering in the planar region and the curvature change region is analyzed to obtain ion exchange distribution information. The residual information of the cleaning solution is compared with the ion exchange distribution information, and the spatially overlapping area where the ion exchange depth is lower than the residual area depth is extracted as the preliminary residual information.

4. The method according to claim 3, characterized in that, The process of constructing the ion exchange distribution information includes: Based on the tempering information set, the stress distribution difference of the irregular curved surface cover plate in the planar area and the curvature change area caused by the different curvature of the surface is analyzed to obtain tempering stress distribution information; Based on the tempering stress distribution information, the modulation direction and amplitude of the stress gradient in each region on the ion diffusion activation energy are analyzed to obtain the ion exchange distribution information.

5. The method according to claim 2, characterized in that, The process involves analyzing the uniformity of ink coverage and masking effect in the initial residue area during screen printing, based on the preliminary residue information and the screen printing information set, to obtain the residue information set, which includes: Based on the preliminary residual information, the local micro-separation state caused by the difference in shrinkage stress during the curing and shrinkage process of the ink is analyzed to obtain the coverage uniformity information; Based on the coverage uniformity information and combined with the screen printing information set, the influence of stress difference between the residual and non-residual areas of the cured ink layer on the optical transmission characteristics is analyzed to obtain masking effect information. The coverage uniformity information and the masking effect information are integrated to construct the residual information set.

6. The method according to claim 5, characterized in that, Based on the residual information set, the secondary ion exchange effect between the low-melting-point glass powder of the ink and the chemically tempered layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect, are analyzed to obtain a defect correlation information set, including: Based on the residual information set, the interfacial contact gap caused by the local micro-separation state of the low-melting-point glass powder in the residual area is analyzed, as well as the contact characteristics formed by stress differences in the non-residual area, to obtain interfacial contact characteristic information. Based on the interface contact characteristic information, the differences in ion exchange activation energy caused by different contact states between the ink layer and the chemical tempering layer in each region during the sintering process are analyzed to obtain secondary ion exchange information; Based on the secondary ion exchange information, the changes in surface layer ion concentration in each region caused by differences in potassium and sodium ion exchange, as well as the resulting redistribution of local compressive stress, are analyzed to obtain the defect association information set.

7. The method according to claim 6, characterized in that, The process of constructing the secondary ion exchange information includes: Based on the interface contact feature information, the thermal expansion behavior of residual gas in the contact gap between the ink layer and the chemical tempering layer in each region during the sintering heating process is analyzed to obtain gap heat transfer difference information. Based on the gap heat transfer difference information, the flow and filling characteristics of low melting point glass powder in the gap caused by the different heat transfer rates in each region are analyzed to obtain melt coverage-contact state information. Based on the melt coverage-contact state information, the modulation effect of the actual ion exchange interface between the low-melting-point glass powder melt and the tempered layer on the ion exchange activation energy is analyzed to obtain the secondary ion exchange information.

8. The method according to claim 6, characterized in that, Based on the secondary ion exchange information, the changes in surface layer ion concentration in each region caused by differences in potassium and sodium ion exchange, and the resulting redistribution of local compressive stress, are analyzed to obtain the defect correlation information set, including: Based on the secondary ion exchange information, the non-uniform accumulation of ion concentration during sintering and cooling process caused by the difference in exchange rates of potassium and sodium ions in the residual and non-residual regions is analyzed to obtain ion concentration gradient evolution information. Based on the ion concentration gradient evolution information, the non-uniform release characteristics of local compressive stress caused by ion migration obstruction during the cooling stage are analyzed to obtain the defect association information set.

9. The method according to claim 8, characterized in that, The step of generating and outputting cover plate defect discrimination information based on the defect association information set includes: Based on the defect association information set, the deviation information between the defect characteristics of each region and the preset cover plate quality standard is analyzed to obtain the defect severity level; Based on the severity level of the defect, generate and output the cover plate defect discrimination information, including the defect location, defect type and defect level.

10. A cover plate defect identification system based on the fusion of cleaning and tempered screen printing data, characterized in that, The method applied to any one of claims 1-9 includes: The watermark residue module is used to acquire the information set of the irregular curved surface cover plate. Based on the information set of the irregular curved surface cover plate, the module analyzes the invisible ion watermark residue information caused by the limitation of the curved surface structure and chemical tempering to obtain the residue information set. The defect association module is used to analyze the secondary ion exchange effect between the low-melting-point glass powder of the ink and the chemical tempering layer during screen printing, and the changes in surface stress distribution caused by the secondary ion exchange effect, based on the residual information set, to obtain the defect association information set. The defect identification module is used to generate and output cover plate defect identification information based on the defect association information set.

Images