Microscopic image-based prepreg resin infiltration uniformity analysis method and system
By combining the adaptive thresholding method and watershed algorithm with Voronoi subdivision technology, the problem of fiber adhesion in microscopic images was solved. Fiber-dominated polygons were constructed, and the permeation resistance coefficient and stress concentration factor were calculated, thus achieving accurate evaluation of the microscopic wetting quality of prepregs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHAANXI HUANGHE XINXING EQUIP CO LTD
- Filing Date
- 2026-01-30
- Publication Date
- 2026-05-19
AI Technical Summary
Existing microscopic image analysis techniques are insufficient to effectively address the visual adhesion problem between reinforcing fiber monofilaments, and traditional wetting evaluation methods cannot adapt to the random topology of fibers, resulting in low accuracy in prepreg quality assessment.
Image segmentation is performed using an adaptive thresholding method and a watershed algorithm. Combined with Voronoi subdivision technology, fiber-dominated polygons are generated, and the permeation resistance coefficient and micro-stress concentration factor are calculated to comprehensively evaluate the wetting reliability probability index.
It achieves nonlinear quantitative grading of the microscopic wetting quality of prepregs, identifies potential defects that are difficult to detect by traditional methods, and improves the accuracy and reliability of the assessment.
Smart Images
Figure CN121600507B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of image processing technology. More specifically, this invention relates to a method and system for analyzing the uniformity of prepreg resin impregnation based on microscopic images. Background Technology
[0002] As a core intermediate material in the field of advanced composite material manufacturing, the microscopic wetting quality of prepregs directly determines the interlaminar shear strength, fatigue resistance, and overall structural safety of the final cured components. In applications with extremely high material performance requirements, such as aerospace and high-end equipment manufacturing, ensuring the full encapsulation and uniform wetting of reinforcing fiber monofilaments by the resin matrix is crucial. Currently, the industry typically uses scanning electron microscopy to photograph the cross-section of cured prepreg samples, obtaining high-resolution microscopic cross-sectional images. Image processing techniques are then used to analyze the distribution of reinforcing fiber monofilaments and the resin matrix, serving as an important basis for evaluating the microscopic process quality of composite materials and predicting macroscopic mechanical properties.
[0003] However, with the increasing demands for fiber volume content in high-performance composite materials, the arrangement of reinforcing fiber monofilaments in the resin matrix is becoming increasingly dense and highly random. In existing microscopic image analysis techniques, it is often difficult to effectively solve the problem of visual adhesion between reinforcing fiber monofilaments when faced with high-density fiber distribution, resulting in the inability to accurately extract the independent geometric boundaries of individual fibers. In addition, traditional wetting evaluation methods often use fixed-size grid division or simple resin area ratio calculation. This rigid spatial discretization method cannot adapt to the natural topological structure of the random aggregation of reinforcing fiber monofilaments. It is easy to misjudge normal small gaps inside dense fiber areas as defects, or simply equate extremely irregular areas with regular areas, failing to objectively reflect the true properties of the microscopic physical space, which seriously affects the accuracy and reliability of prepreg quality assessment. Summary of the Invention
[0004] To address the technical problems of existing technologies, such as the difficulty in solving the problem of fiber visual adhesion segmentation and the inability of rigid spatial discreteness to adapt to random topological structures, resulting in low evaluation accuracy, this invention provides solutions in the following aspects.
[0005] In a first aspect, the present invention provides a method for analyzing the uniformity of prepreg resin impregnation based on microscopic images, comprising:
[0006] A microscopic cross-sectional image of the prepreg sample to be tested is acquired. The microscopic cross-sectional image is segmented, and individual reinforcing fiber monofilaments are obtained based on the segmentation results. The centroid coordinates of the reinforcing fiber monofilaments are extracted, and Voronoi diagrams are performed on the two-dimensional plane space where the microscopic cross-sectional image is located based on the centroid coordinates to generate fiber domain polygons corresponding to each reinforcing fiber monofilament. The permeation resistance coefficient of the reinforcing fiber monofilament is calculated based on the cross-sectional area of the reinforcing fiber monofilament, the total area of the fiber domain polygons, and the total perimeter. The microscopic stress concentration factor of the reinforcing fiber monofilament is calculated based on the minimum wall distance from the centroid of the reinforcing fiber monofilament to the boundary of the fiber domain polygon and the equivalent radius of the reinforcing fiber monofilament. The wetting reliability probability index corresponding to the reinforcing fiber monofilament is calculated based on the permeation resistance coefficient and the microscopic stress concentration factor. The wetting quality level of the region to which the reinforcing fiber monofilament belongs is determined based on the magnitude of the wetting reliability probability index.
[0007] This invention, by segmenting microscopic cross-sectional images, can accurately extract the independent morphology of reinforcing fiber monofilaments from complex microscopic backgrounds. It then utilizes Voronoi segmentation technology to generate physical domains that conform to the natural arrangement of the fibers, thus establishing an objective spatial analysis benchmark for irregular microstructures. Based on this, the invention calculates the permeation resistance coefficient and micro-stress concentration factor to measure the viscous resistance of resin flow from a fluid dynamics perspective and assess the stress singularity risk caused by extreme fiber proximity from a solid mechanics perspective. Finally, these two physical dimensions are coupled into a wetting reliability probability index, achieving a comprehensive quantitative classification of the micro-wetting quality of prepregs. This effectively identifies potential defects caused by channel blockage or stress concentration that are difficult to detect using traditional methods, providing more scientific data support for optimizing composite material manufacturing processes.
[0008] Preferably, the step of segmenting the microscopic cross-sectional image and obtaining independent reinforcing fiber filaments based on the segmentation results includes: performing binarization processing on the microscopic cross-sectional image using an adaptive thresholding method to obtain a binarized image; calculating the Euclidean distance from each foreground pixel to the nearest background pixel in the binarized image to obtain a distance transformation map; extracting local maxima points in the distance transformation map as seed points; using the watershed algorithm to expand outward from the seed points to segment the adhered region into multiple independent connected domains, and treating each connected domain as a reinforcing fiber filament.
[0009] This invention employs an adaptive thresholding method combined with distance transformation and a watershed algorithm to segment images. By calculating the distance from the foreground pixel to the background and extracting local maxima, the geometric center of each filament within the adhesion region can be accurately located. The watershed algorithm is then used to expand outward from these centers following the topological terrain, thereby constructing precise segmentation boundaries between physically closely contacting but logically independent fibers. This effectively solves the common visual adhesion problem in high volume fraction prepreg microscopy and ensures the accuracy of subsequent independent geometric analysis of each reinforcing fiber filament.
[0010] Preferably, the step of performing Voronoi subdivision on the two-dimensional plane space where the microscopic cross-sectional image is located based on the centroid coordinates to generate the fiber domain polygon corresponding to each reinforcing fiber monofilament includes: using the centroid coordinates of all reinforcing fiber monofilaments as seed point set, performing Voronoi subdivision on the two-dimensional plane space where the microscopic cross-sectional image is located; for regions located at the edge of the microscopic cross-sectional image, using the image bounding box for truncation and closure processing to generate multiple closed polygon regions, and defining each closed polygon region as the fiber domain polygon of the corresponding reinforcing fiber monofilament.
[0011] This invention utilizes the centroid coordinates of reinforcing fiber monofilaments to perform Voronoi subdivision and edge closure processing on a two-dimensional planar space. Based on the principle of proximity, the continuous resin matrix space can be objectively discretized into the physical domain exclusively occupied by each reinforcing fiber monofilament. This division method conforms to the natural topological structure of the random arrangement of fibers and avoids the fragmentation of irregular microstructures caused by the traditional fixed grid method. Thus, it provides a spatial benchmark with clear geometric and physical significance for evaluating the resin distribution around a single fiber.
[0012] Preferably, the permeation resistance coefficient satisfies the expression: ;in, This indicates the permeation resistance coefficient of the reinforcing fiber monofilament; This indicates the cross-sectional area of the reinforcing fiber monofilament; Represents the total area of the polygon within the fiber's domain; Pi; This represents the total perimeter of the polygon within the fiber's domain; It is an exponential function with the natural constant as its base.
[0013] The permeation resistance coefficient constructed in this invention takes into account both the spatial occupancy ratio of the reinforcing fiber monofilament within the local domain and the regularity of the domain shape. By nonlinearly amplifying the shape distortion through an exponential function, it can keenly capture the huge viscous resistance generated when fiber agglomeration causes the resin channel to become narrow or twisted. Thus, it measures the difficulty of resin matrix flow and permeation in micropores from a fluid dynamics perspective and can effectively identify potential poor wetting areas caused by channel blockage or distortion.
[0014] Preferably, the method for obtaining the minimum wall distance is as follows: obtain the vertical distance from the centroid coordinates of the reinforcing fiber monofilament to each side of its corresponding fiber domain polygon, and take the minimum value of the vertical distance as the minimum wall distance.
[0015] Preferably, the equivalent radius satisfies the expression: In the formula, This indicates the equivalent radius of the reinforcing fiber monofilament; This indicates the cross-sectional area of the reinforcing fiber monofilament; Pi is the mathematical constant of a circle.
[0016] Preferably, the micro-stress concentration factor satisfies the expression: ;in, This represents the micro-stress concentration factor of the reinforcing fiber monofilament; This indicates the equivalent radius of the reinforcing fiber monofilament; Indicates the minimum wall distance; It is a natural constant; It is the natural logarithm function.
[0017] This invention calculates the micro-stress concentration factor by obtaining the minimum wall distance from the centroid of the reinforcing fiber monofilament to the boundary of the domain and combining it with the equivalent radius. This can reflect the singularity abrupt change in the stress field caused by the extreme proximity of fibers in composite material mechanics. When the fiber spacing approaches zero, the micro-stress concentration factor can rapidly increase in a trend that conforms to physical laws. This effectively characterizes the stress concentration risk caused by the excessive thinness of the resin layer and helps the detection system to locate structural weak points that are prone to microcracks during the curing process in advance.
[0018] Preferably, the wetting reliability probability index satisfies the expression: ;in, This represents the probability index of wetting reliability corresponding to the reinforcing fiber monofilament; This indicates the permeation resistance coefficient of the reinforcing fiber monofilament; This represents the micro-stress concentration factor of the reinforcing fiber monofilament.
[0019] This invention couples the permeation resistance coefficient, which characterizes the difficulty of fluid permeation, with the microscopic stress concentration factor, which characterizes the mechanical risk of solids, into a normalized wetting reliability probability index. It adopts a parallel failure evaluation logic, that is, whenever the flow resistance is too large and affects wetting or the stress is too large and leads to the risk of cracking, the wetting reliability probability index will respond sensitively. This enables a comprehensive evaluation of the microscopic quality of prepreg from two dimensions: manufacturing processability and service reliability, and provides a rigorous decision basis for the final quality classification.
[0020] Preferably, determining the wetting quality level of the region to which the reinforcing fiber monofilament belongs based on the wetting reliability probability index includes: determining the region to which the reinforcing fiber monofilament belongs as a high-quality wetting region in response to the wetting reliability probability index being greater than or equal to a preset safety threshold; determining the region to which the reinforcing fiber monofilament belongs as a transition warning region in response to the wetting reliability probability index being less than the preset safety threshold but greater than or equal to a preset critical threshold; and determining the region to which the reinforcing fiber monofilament belongs as a defective region in response to the wetting reliability probability index being less than the preset critical threshold.
[0021] Secondly, the present invention provides a prepreg resin wetting uniformity analysis system based on microscopic images, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned prepreg resin wetting uniformity analysis method based on microscopic images is implemented.
[0022] By adopting the above technical solution, the above-mentioned method for analyzing the uniformity of prepreg resin impregnation based on microscopic images is generated into a computer program and stored in a memory so that it can be loaded and executed by a processor. In this way, a terminal device can be made based on the memory and the processor for convenient use.
[0023] The beneficial effects of this invention are as follows: By adaptively segmenting microscopic cross-sectional images, this invention can identify the independent morphology of reinforcing fiber monofilaments and construct fiber-dominated polygons that conform to the natural random arrangement of fibers using Voronoi diagrams, overcoming the shortcomings of traditional methods in handling fiber adhesion and unreasonable fixed mesh division. On this basis, this invention comprehensively considers fluid dynamics and solid mechanics mechanisms, calculates the permeation resistance coefficient based on the degree of geometric distortion of the fiber-dominated polygons to characterize the viscous resistance of resin flow, and calculates the micro-stress concentration factor based on the distance between the reinforcing fiber monofilament and its nearest neighbor to assess the risk of curing cracking. These two dimensions of physical indicators are coupled into a unified wetting reliability probability index, thereby realizing nonlinear quantitative grading of the micro-wetting quality of prepregs. This can effectively detect potential hidden defects caused by micro-channel blockage or extremely close proximity of fibers, improving the accuracy of prepreg quality assessment. Attached Figure Description
[0024] Figure 1 This is a flowchart illustrating the method for analyzing the uniformity of prepreg resin wetting based on microscopic images in this invention.
[0025] Figure 2 This is a schematic diagram of a microscopic cross-sectional image in an embodiment of the present invention;
[0026] Figure 3 This is a schematic diagram illustrating the reinforcement of fiber monofilament segmentation and fiber domain polygon construction in an embodiment of the present invention;
[0027] Figure 4 This is a scatter plot showing the correlation between the permeation resistance coefficient and the micro-stress concentration factor in an embodiment of the present invention.
[0028] Figure 5 This is a heatmap showing the distribution of the wetting reliability probability index in an embodiment of the present invention. Detailed Implementation
[0029] The technical solutions of the embodiments of the present invention 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 the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0030] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0031] This invention discloses a method for analyzing the uniformity of prepreg resin impregnation based on microscopic images, referring to... Figure 1 This includes steps S1-S5:
[0032] S1. Obtain a microscopic cross-sectional image of the prepreg sample to be tested, perform image segmentation on the microscopic cross-sectional image, and obtain independent reinforcing fiber monofilaments based on the segmentation results.
[0033] It should be noted that in the microscopic quality analysis of prepregs, the diameter of the reinforcing fiber monofilaments is typically extremely small, and the resin matrix is fluid in its uncured state. This makes the grayscale difference between the fiber and the resin easily affected by thermal noise during imaging. Furthermore, the close arrangement of fibers in physical space can lead to visual adhesion in microscopic imaging. If the independent boundaries of individual fibers cannot be accurately identified, subsequent quantitative analysis of the fiber's microscopic topology will lack a reliable geometric reference, resulting in distorted wetting quality assessment results. Therefore, this invention uses physical sample preparation combined with adaptive photoelectric signal processing logic to transform the material's microscopic physical phases into a high-precision digital feature matrix.
[0034] Specifically, the prepreg sample to be tested is subjected to resin embedding, curing, grinding, and polishing to prepare a metallographic observation sample. A scanning electron microscope is used in backscattered electron mode to image the cross-section of the metallographic observation sample, obtaining a microscopic cross-sectional image. For example, Figure 2 This is a schematic diagram of a microscopic cross-sectional image in an embodiment of the present invention.
[0035] Gaussian filtering was applied to the microscopic cross-sectional image, and the filtered image was binarized using an adaptive thresholding method. The image was then segmented into a foreground pixel set and a background pixel set. The foreground pixel set represents the reinforcing fiber monofilament, and the background pixel set represents the resin matrix.
[0036] Furthermore, for the fiber adhesion regions present in the binarized image, a distance transform combined with a watershed algorithm is used for separation. Specifically: the Euclidean distance from each foreground pixel to the nearest background pixel is calculated to obtain a distance transform map. Local maxima in the distance transform map are extracted as seed points. The watershed algorithm is then used to expand outward from the seed points to segment the distance transform map, resulting in multiple connected components. Each connected component is treated as a reinforcing fiber filament. Each reinforcing fiber filament is traversed, and the centroid coordinates, cross-sectional area, and perimeter of the reinforcing fiber filament are extracted through connected component pixel statistics.
[0037] S2. Extract the centroid coordinates of the reinforcing fiber monofilaments, and perform Voronoi subdivision on the two-dimensional plane space where the microscopic cross-sectional image is located based on the centroid coordinates to generate the fiber domain polygon corresponding to each reinforcing fiber monofilament.
[0038] It should be noted that, in the actual layup process, the reinforcing fiber monofilaments inside the prepreg exhibit physically unavoidable random agglomeration characteristics due to the influence of tension and resin rheology. Existing fixed-grid analysis methods forcibly map irregular fiber arrangements onto regular rectangular grids. When the grid size is not set properly, it is easy to misjudge normal micro-gaps within fiber agglomeration areas as poor wetting, or to miss micro-bubbles in resin-rich areas. To objectively characterize the resin range actually controlled by each fiber in physical space, this invention utilizes the Voronoi geometric segmentation characteristics to discretize the continuous resin matrix space into a physical domain exclusively for each fiber based on the natural arrangement of the reinforcing fiber monofilaments.
[0039] Specifically, the centroid coordinates of all the reinforcing fiber monofilaments are used as seed points to perform Voronoi subdivision on the two-dimensional plane space containing the microscopic cross-sectional image. For regions located at the edge of the microscopic cross-sectional image, image bounding boxes are used for truncation and closure to ensure that all generated regions are closed shapes. Each closed polygon region generated by the subdivision is defined as the fiber domain polygon of the corresponding fiber.
[0040] Furthermore, for each reinforcing fiber monofilament, its associated fiber domain polygon is analyzed, and the geometric centroid, total area, and total perimeter of the fiber domain polygon are extracted. The total area of the fiber domain polygon includes the area of the reinforcing fiber monofilament itself and the area of the resin matrix surrounding the fiber.
[0041] For example, Figure 3 This is a schematic diagram illustrating the reinforcement of fiber monofilament segmentation and fiber domain polygon construction in an embodiment of the present invention. Figure 3 In the fiber agglomeration area in the lower right corner, the grid lines become extremely dense and adhere closely to the fiber surface, reflecting the physical characteristics of closed resin channels in this area; while in the sparse area in the upper left corner, the grid lines are wide and extended, reflecting the physical characteristics of abundant resin in this area.
[0042] S3. Calculate the permeation resistance coefficient of the reinforcing fiber monofilament based on the cross-sectional area of the reinforcing fiber monofilament, the total area of the fiber-dominated polygon, and the total perimeter.
[0043] It should be noted that traditional wetting analysis only focuses on the absolute value of resin area, ignoring the restrictive effect of microchannel shape on fluid permeation. According to the principles of microfluidic dynamics, the permeation efficiency of resin in porous media is closely related to the geometric shape factor of the pore channels. In regions where reinforcing fiber monofilaments are severely aggregated, the fiber-dominated polygons often exhibit extremely flat or distorted shapes. This shape generates enormous viscous resistance, making it difficult for the resin to flow effectively and encapsulate the fibers. Simple area ratio calculations cannot reveal this wetting failure caused by high flow resistance. Therefore, this invention introduces a shape factor based on the isoperimeter quotient to construct a permeation resistance coefficient, measuring the kinetic resistance of micro-permeation.
[0044] Specifically, the permeation resistance coefficient of the reinforcing fiber monofilament is calculated based on the cross-sectional area of the reinforcing fiber monofilament, the total area of the fiber-dominated polygon, and the total perimeter of the fiber-dominated polygon.
[0045] The permeation resistance coefficient satisfies the expression:
[0046]
[0047] in, This indicates the permeation resistance coefficient of the reinforcing fiber monofilament; This indicates the cross-sectional area of the reinforcing fiber monofilament; Represents the total area of the polygon within the fiber's domain; This represents the area of the pure resin matrix within the fiber-dominated polygon. It characterizes the space occupancy ratio of the reinforcing fiber monofilament relative to the resin matrix, i.e. the degree of crowding of the micro-channels. When the area of the pure resin matrix approaches 0, this term tends to infinity, reflecting the ultimate cutoff state when the micro-fluid channels are completely squeezed and closed. Pi; This represents the total perimeter of the polygon within the fiber's domain; It is the reciprocal form of the isocircular quotient, used to characterize the non-circularity of the fiber-dominated polygon; It is an exponential function with the natural constant as its base. When the reinforcing fiber monofilaments are uniformly distributed, the fiber domain polygon is close to a regular polygon or a circle, and the isopecile quotient is... Approaching 1 makes the exponent term close to 1. The permeation resistance coefficient is close to 1. Lower; when the reinforcing fiber monofilaments agglomerate, causing the fiber domain polygon to become a narrow strip or an irregular polygon, Much larger This leads to the exponential term It increases exponentially, causing the permeation resistance coefficient to... The significant increase reflects the enormous viscous drag generated by the shape distortion.
[0048] S4. Calculate the micro-stress concentration factor of the reinforcing fiber monofilament based on the minimum wall distance from the centroid of the reinforcing fiber monofilament to the boundary of the fiber-dominated polygon and the equivalent radius of the reinforcing fiber monofilament.
[0049] It should be noted that the linear offset of the fiber centroid relative to the local centroid cannot fully reflect the true mechanical risk. In the fracture mechanics of composite materials, microcracks often initiate at the weakest point where the fiber spacing is smallest. Even if the centroid offset of the reinforcing fiber monofilament is small, if its edge is extremely close to the neighboring reinforcing fiber monofilament, the resin layer at that point will be extremely thin, resulting in an extremely high stress concentration factor after curing. Therefore, this invention constructs a micro-stress concentration factor based on the nearest neighbor wall distance to assess the micro-mechanical risk caused by the geometric proximity effect.
[0050] Specifically, obtain the perpendicular distances from the centroid of the reinforcing fiber monofilament to each edge of its corresponding fiber domain polygon, and record the minimum value as the minimum wall distance; simultaneously calculate the equivalent radius of the reinforcing fiber monofilament:
[0051]
[0052] In the formula, This indicates the equivalent radius of the reinforcing fiber monofilament; This indicates the cross-sectional area of the reinforcing fiber monofilament; Pi is the mathematical constant of a circle.
[0053] Furthermore, based on the minimum wall distance and equivalent radius, the micro-stress concentration factor corresponding to the reinforcing fiber monofilament is calculated.
[0054] The micro-stress concentration factor satisfies the following expression:
[0055]
[0056] in, The micro-stress concentration factor represents the degree of stress concentration between the reinforcing fiber monofilament and its nearest neighbor due to their close proximity. This indicates the equivalent radius of the reinforcing fiber monofilament; This represents the minimum wall distance. Due to the characteristics of Voronoi splitting, the minimum wall distance is equal to half the distance between the centroid of the reinforcing fiber monofilament and the nearest neighboring reinforcing fiber monofilament. It is a natural constant; It is a natural logarithmic function. When the reinforcing fiber monofilaments are uniformly distributed in the resin matrix and do not interfere with each other, Larger, making Smaller, micro-stress concentration factor Maintaining a lower baseline level reflects sufficient resin layer thickness between fibers and smooth stress transmission in this region; as the reinforcing fiber monofilaments gradually approach their neighboring fibers, i.e., they tend to aggregate or touch, the resin layer between fibers becomes thinner. The value decreases rapidly and approaches 0, at which point the ratio... The rapid increase leads to a micro-stress concentration factor The rapid increase reflects a singular abrupt change in the stress field in the narrow slit region extremely close to the fibers, making the matrix highly susceptible to cracking due to stress concentration. This invention introduces a natural constant. and the natural logarithm function This is to simulate the physical law that stress concentration effects deteriorate nonlinearly with decreasing distance, while ensuring that when When the value is large, the micro-stress concentration factor has a stable basic value.
[0057] S5. Based on the permeation resistance coefficient and the micro-stress concentration factor, calculate the wetting reliability probability index corresponding to the reinforcing fiber monofilament, and determine the wetting quality level of the region to which the reinforcing fiber monofilament belongs based on the magnitude of the wetting reliability probability index.
[0058] It should be noted that a high-quality prepreg microstructure must simultaneously meet two conditions: low flow resistance to ensure resin penetration and low stress to prevent curing cracking. This invention couples the flow resistance coefficient, which characterizes fluid dynamics, with the micro-stress concentration factor, which characterizes solid mechanics, into a normalized comprehensive index, thereby achieving nonlinear grading of wetting quality.
[0059] Specifically, based on the permeation resistance coefficient and the micro-stress concentration factor, the wetting reliability probability index corresponding to each reinforcing fiber monofilament is calculated.
[0060] The wetting reliability probability index satisfies the expression:
[0061]
[0062] in, This represents the probability index of wetting reliability corresponding to the reinforcing fiber monofilament; This indicates the permeation resistance coefficient of the reinforcing fiber monofilament; This represents the micro-stress concentration factor of the reinforcing fiber monofilament. This invention employs the physical logic of a parallel failure model, which... and Considered as two negative loading factors that disrupt system stability, when the permeation resistance coefficient... A large value indicates that the resin is difficult to flow, or a micro-stress concentration factor. When the denominator is very large, indicating a risk of cracking, The rapid increase leads to a higher probability index of wettability. A rapid approach to 0 indicates a defect in the region of the reinforcing fiber monofilament; only when the permeation resistance coefficient... Smaller and micro-stress concentration factor When the denominator is small, Approaching 1, wetting reliability probability index A value close to 1 indicates that the wetting quality of the monofilament region of the reinforcing fiber is excellent.
[0063] For example, Figure 4 This is a scatter plot showing the correlation between the permeation resistance coefficient and the micro-stress concentration factor in an embodiment of the present invention. Figure 4 It can be seen that the wetting reliability probability index is only high when both the permeation resistance coefficient and the micro-stress concentration factor are in the dense region in the lower left corner of the coordinate axis. Once either the permeation resistance coefficient or the micro-stress concentration factor deteriorates, causing the data points to diverge to the right or upward, the wetting reliability probability index drops rapidly, thus verifying the invention's ability to keenly capture high flow resistance and high stress defects.
[0064] Figure 5 This is a heatmap showing the distribution of the wetting reliability probability index in an embodiment of the present invention. Figure 5As can be seen, the present invention can clearly distinguish and visualize the differences in microscopic wetting quality in the prepreg, especially the low-value clumps in the lower right corner, which accurately locate the wetting failure area caused by severe fiber agglomeration.
[0065] Furthermore, quality grading thresholds are set, including safety thresholds. and critical threshold The wetting reliability probability index of each reinforcing fiber monofilament is compared with the quality grading threshold to determine its quality level. Specifically:
[0066] In response to The region to which the reinforcing fiber monofilament belongs is determined to be a high-quality impregnation zone; in response to The region to which the reinforcing fiber monofilament belongs is determined to be a critical transition warning zone for wetting; in response to The area to which the reinforcing fiber monofilament belongs is determined to be a defective area with poor wetting.
[0067] It should be noted that, in this embodiment, a safety threshold is set. Its setting logic is that when the immersion reliability probability index Greater than or equal to the safety threshold hour, This means that the negative load factor of the system is less than one-quarter of the basic stability constant 1. At this point, the system is in a micro-perturbation steady state, and the effects of flow resistance and stress can be regarded as fluctuations within the elastic range, which will not have a substantial impact on the final curing quality. Therefore, it is defined as a high-quality wetting zone; a critical threshold is set. Its setting logic is that when the immersion reliability probability index Less than the critical threshold hour, When this threshold is reached, it means that the negative load factor of the system has exceeded 1.5 times the basic stability constant 1. At this point, the system enters a defect-dominated state, where excessive flow resistance or stress concentration becomes the decisive factor, greatly increasing the probability of a decrease in interlaminar shear strength or the propagation of microcracks. Therefore, this is defined as a defect zone. In other embodiments, implementers can use the standard sample statistical calibration method to set the above threshold: select a set of standard prepreg samples that have been verified as qualified through destructive testing, and calculate the mean value of the wetting reliability probability index of all reinforcing fiber monofilaments in the entire sample. and standard deviation Set security threshold Set a critical threshold This allows for the establishment of dynamic evaluation standards that conform to specific process tolerance zones.
[0068] This invention also discloses a microscopic image-based prepreg resin wetting uniformity analysis system, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the microscopic image-based prepreg resin wetting uniformity analysis method according to this invention.
[0069] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.
Claims
1. A method for analyzing the uniformity of prepreg resin impregnation based on microscopic images, characterized in that, include: Acquire a microscopic cross-sectional image of the prepreg sample to be tested, perform image segmentation on the microscopic cross-sectional image, and obtain independent reinforcing fiber monofilaments based on the segmentation results; The centroid coordinates of the reinforcing fiber monofilaments are extracted, and the two-dimensional plane space where the microscopic cross-sectional image is located is divided into Voronoi sections based on the centroid coordinates to generate the fiber domain polygon corresponding to each reinforcing fiber monofilament. Calculate the permeation resistance coefficient of the reinforcing fiber monofilament based on its cross-sectional area, the total area of the polygon within the fiber domain, and the total perimeter. ;in, This represents the permeation resistance coefficient of the reinforcing fiber monofilament; This indicates the cross-sectional area of the reinforcing fiber monofilament; This represents the total area of the polygon within the fiber's domain; Pi; This represents the total perimeter of the polygon within the fiber's domain; It is an exponential function with the natural constant as its base; The micro-stress concentration factor of the reinforcing fiber monofilament is calculated based on the minimum wall distance from the centroid of the reinforcing fiber monofilament to the boundary of the fiber-dominated polygon, and the equivalent radius of the reinforcing fiber monofilament. ;in, This represents the micro-stress concentration factor of the reinforcing fiber monofilament; This indicates the equivalent radius of the reinforcing fiber monofilament; Indicates the minimum wall distance; It is a natural constant; It is the natural logarithm function; Based on the permeation resistance coefficient and the micro-stress concentration factor, the wetting reliability probability index corresponding to the reinforcing fiber monofilament is calculated. ;in, This indicates the wetting reliability probability index corresponding to the reinforcing fiber monofilament; based on the magnitude of the wetting reliability probability index, the wetting quality level of the region to which the reinforcing fiber monofilament belongs is determined.
2. The method for analyzing the uniformity of prepreg resin impregnation based on microscopic images according to claim 1, characterized in that, The step of segmenting the microscopic cross-sectional image and obtaining independent reinforcing fiber monofilaments based on the segmentation results includes: An adaptive thresholding method was used to binarize the microscopic cross-sectional image to obtain a binarized image. The Euclidean distance from each foreground pixel to the nearest background pixel in the binarized image was calculated to obtain a distance transformation map. Local maxima in the distance transformation map were extracted as seed points. The watershed algorithm was used to expand outward from the seed points to divide the adhered region into multiple independent connected domains. Each connected domain was used as a reinforcing fiber filament.
3. The method for analyzing the uniformity of prepreg resin impregnation based on microscopic images according to claim 1, characterized in that, The Voronoi subdivision of the two-dimensional plane space containing the microscopic cross-sectional image based on centroid coordinates, generating the fiber domain polygon corresponding to each reinforcing fiber monofilament, includes: Using the centroid coordinates of all reinforcing fiber monofilaments as seed point sets, the two-dimensional plane space where the microscopic cross-sectional image is located is divided by Voronoi subdivision; for the region located at the edge of the microscopic cross-sectional image, the image bounding box is used to perform truncation and closure processing to generate multiple closed polygonal regions, and each closed polygonal region is defined as the fiber domain polygon of the corresponding reinforcing fiber monofilament.
4. The method for analyzing the uniformity of prepreg resin impregnation based on microscopic images according to claim 1, characterized in that, The method for obtaining the minimum wall distance is as follows: Obtain the perpendicular distances from the centroid coordinates of the reinforcing fiber monofilament to each edge of its corresponding fiber domain polygon, and take the minimum value among the perpendicular distances as the minimum wall distance.
5. The method for analyzing the uniformity of prepreg resin impregnation based on microscopic images according to claim 1, characterized in that, The equivalent radius satisfies the expression: ； In the formula, This indicates the equivalent radius of the reinforcing fiber monofilament; This indicates the cross-sectional area of the reinforcing fiber monofilament; Pi is the mathematical constant of a circle.
6. The method for analyzing the uniformity of prepreg resin impregnation based on microscopic images according to claim 1, characterized in that, The determination of the wetting quality level of the region to which the reinforcing fiber monofilament belongs based on the magnitude of the wetting reliability probability index includes: When the wetting reliability probability index is greater than or equal to a preset safety threshold, the area to which the reinforcing fiber monofilament belongs is determined to be a high-quality wetting area; when the wetting reliability probability index is less than the preset safety threshold but greater than or equal to a preset critical threshold, the area to which the reinforcing fiber monofilament belongs is determined to be a transitional warning area; when the wetting reliability probability index is less than the preset critical threshold, the area to which the reinforcing fiber monofilament belongs is determined to be a defective area.
7. A microscopic image-based system for analyzing the uniformity of prepreg resin impregnation, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the method for analyzing the uniformity of prepreg impregnation based on microscopic images according to any one of claims 1-6.