Model Reconstruction And Mechanical Properties Of Carbon Black Filled Rubber Network Based On Electron Microscope Images

Carbon black as an ideal filler is widely used to enhance the mechanical properties of rubber.When studying the microscopic morphology modeling of real carbon black,the existing compound rubber-based modeling method is affected by the true morphology of the particles.At the same time,there are defects in the carbon black particle stacking display in the composite rubber electron microscope image.It is easy to cause large errors in impact.The three-dimensional scanning method based on electron microscopy images generally uses costly machine facilities and relatively simple image processing methods for three-dimensional reconstruction.There is room for improvement in the accuracy and methods of image processing and analysis of carbon black morphology.This paper uses image stacking technology for model reconstruction and analyzes the feasibility of the reconstruction method;researches on background processing and fitting algorithms for carbon black morphological image processing,and objective evaluation of the algorithm through image indicators;using fitting images Reconstruction of the two-dimensional network model for the real electron microscope image and reconstruction of the three-dimensional model based on the assumption that the carbon black particles in the image are located on the central axis,and the mechanical stretching simulation calculation of the two-dimensional network reconstruction model;using the ideal network model for Payne The relationship between the effect and the filler network is simulated and analyzed.The specific research and results are as follows:(1)The feasibility of the stacked image reconstruction model method was studied.Firstly,a representative volume unit(RVE)model with uniform distribution of carbon black is established.Based on the image stacking technology,one hundred original model cross-sectional pictures are stacked to form a reconstruction model.The main factors affecting the accuracy of the reconstruction model are the number of particles,the number of volumes,and the grid.In terms of three aspects,through the establishment of three factors and three levels of orthogonal experimental research,the results show that the order of the three factors on the degree of error of the reconstruction model is:volume fraction>particle number>grid number.Finally,an aggregate model was established for further verification,and it was found that the aggregate model had a smaller error than the original model during the stack modeling process.(2)The defects of uneven light,noise interference,and indistinguish ability of the filler matrix in the TEM electron microscope image were treated with image processing methods.It mainly includes image segmentation of background processing,filtering method and threshold iteration method to distinguish matrix filler.The matrix fillers in the electron microscopy images obtained after processing are clearly distinguished,which lays a foundation for the fitting of subsequent research.(3)The establishment process of the two-dimensional carbon black network model was studied.Firstly,in view of the lack of quantitative description and simulation methods for microscopic electron micrographs of filled rubber composites,three image fitting algorithms based on image morphology and grayscale information are proposed:contour-based contour skeleton algorithm and maximum inscribed circle algorithm,and grayscale-based Valued K-means algorithm,and evaluated the fitting effect of the three algorithms by the peak signal-to-noise ratio(PSNR)and structural similarity(SSIM)image standards.The results show that the contour skeleton algorithm has the best fitting effect on carbon black aggregates.Then the contour skeleton algorithm is used to fit the electron microscope image to obtain a two-dimensional carbon black network reconstruction model.Finally,the ideal two-dimensional network model is established through the same carbon black area ratio as the two-dimensional reconstruction network model,which provides a basis for the study of filler networks in many aspects.(4)The mechanical properties in the network model are studied.The two-dimensional network reconstruction model and the two-dimensional ideal network model were subjected to static mechanical stretching.The stress and deformation degree of the carbon black network and the rubber matrix were studied through the Mises stress cloud diagram and the aspect ratio.The higher stresses were mainly distributed in In the carbon black network,the lower stress is mainly distributed in the rubber matrix,so the carbon black network bears the main stress during the stretching process.The influence of the carbon black filler network on the Payne effect was studied through the network connectivity rate and the network connection direction angle.Compared with the network connection rate,the connection direction angle has a greater influence on the Payne effect.(5)Based on the assumption that all carbon black particles in the image are on the central axis,the 3D model was reconstructed.The stress-strain curve of the 3D model was compared with the 2D reconstruction model curve and experimental data.The stress of the 3D stress-strain curve was slightly higher than that of the 2D model.Compared with the two-dimensional model,the stress-strain curve of the three-dimensional model is closer to the experimental curve.This reconstruction method can be used as a reference for studying 3D reconstruction.