دانلود رایگان مقاله بازسازی سه بعدی ساختار متخلخل سنگ با استفاده از الگوریتم فراکتال

Abstract

Natural rock, such as sandstone, has a large number of discontinuous, multi-scale, geometry-irregular pores, forming a complex porous structure. This porous structure essentially determines the rock’s physical and/or mechanical properties, which are of great significance to a variety of applications in the fields of science and engineering. As a supplement to experimental observation, a reliable reconstruction model of porous structure could provide an effective and economical way to characterize the physical and mechanical properties of a porous rock. In this paper, we present a novel method for reconstructing the well-connected porous structure of sandstones, which are often intractable to handle for current reconstruction methods. A fractal descriptor is here proposed for better characterizing complex pore morphologies. The reconstruction procedure of a 3D well-connected porous structure is optimized by integrating the improved simulated annealing algorithm and the fractal system control function. The proposed reconstruction method enables us to represent a large-size 3D porous structure. To verify the accuracy of reconstruction, we have analyzed the geometrical, topological, and mechanical properties of the reconstructed porous medium and compared them with those of prototype rock samples. The comparisons show good agreement between the reconstructed model and the real porous sandstone.

5. Conclusions

In this paper, a faster, more accurate, and more efficient reconstruction algorithm (compared to current reconstruction algorithms) has been developed for representing the 3D porous microstructure of well-connected sandstone. Specifically, a MIP procedure was incorporated to accelerate the evolution of the preliminary structure, which ensures better final reconstruction results. A fractal descriptor and a fractal system control function were introduced in the reconstruction procedure. The fractal system control function along with the two-point probability control function and the linear-path control function can provide a better description of complex pore structures as well as ensure the effectiveness of reconstruction. Pre-conditioning procedure was carried out to speed up the convergence of reconstruction process. All of these improvements endow the proposed method with the capability of representing the microstructures of rocks with larger sizes and more complex porous structures compared to conventional methods such as the SAA algorithm.