Research On Cement Microelements Modeling Based On Moving Least Squares Surface

As an important building base material,cement plays an important role in the people's social production and life.China is the largest producer in the cement industry,but the quality of cement has not reached the world average level.Therefore,how to combine in-depth research on cement with scientific methods to break through the existing bottleneck of cement quality has become an important issue in cement development.However,the cement hydration process is extremely complicated,and the physical changes and chemical reactions therein are difficult to understand.The microstructure of cement reflects microscopic changes and determines macroscopic properties.So,the modeling of cement microstructure plays an important role in understanding the hydration of cement,and the study of microstructure plays a role in guiding the development of new materials.Advanced image acquisition technology makes the acquisition of cement microstructure images accurate to the nanometer level,which has promoted the rapid development of computer modeling research based on cement image data and intelligent algorithms.Advanced technology can analyze the microstructure of cement to obtain the unknown characteristics of the cement hydration process.However,people have not been able to give a complete and unified mathematical mapping from the real cement microstructure images obtained.And they have not yet solved the highly realistic micro element appearance description method and display technology in computer simulation.Therefore,the research on high-precision fitting of real cement microstructure has become an important research topic to improve the simulation quality of cement microstructure and promote the scientific development of modeling and simulation of cement hydration.The main research contents of this article include the following aspects:(1)Edge detection for cement images based on interactive genetic algorithmAiming at the problems that the cement microstructure image is highly complex,the cement material is isotropic,and the cement edge and internal cement particles are difficult to detect,this paper proposes an edge detection method based on interactive genetic algorithm.Guided by human-computer interaction,accurate edge extraction of complex cement microstructure images is achieved.Experimental results show that the algorithm is effective for cement micro-element images.(2)Surface reconstruction of cement microstructure based on moving least squaresAiming at the problem of the complex point cloud of the cement microstructure image and the difficulty of fitting the point cloud,a new method for reconstructing the three-dimensional surface of the micro-morphology of cement hydrate was proposed.This paper explored the advantages and functions of moving least squares in the fitting of complex microstructure images of cement,and performed fitting experiments on complex point cloud data of cement.In addition,this paper also explored the impact of moving least squares internal parameters,including the selection of the basis function and the size of the tightly supported domains,and the analysis of the basis function selection and the size of the tightly supported domains on the microstructural images of cement influences.(3)Cement microelement fitting of moving least squares surfaces with boundary pointsAiming at the problem of large point cloud and low fitting efficiency in cement surface reconstruction,the paper explored the method of selecting image feature point cloud.A fast feature point cloud selection method for cement microstructure images is proposed.This method introduces a method of combining boundary points with internal feature points.Operate the image by constructing edge response neighborhoods and select important feature points in the edge information.Then use the minimum moving surface to perform microstructure interpolation.A number of experimental analyses have been carried out.Experimental results show that this method can achieve rapid reconstruction of cement microstructures while ensuring accuracy.