微晶玻璃晶相生长过程的Monte Carlo模拟

Computer simulations of crystalline grain growth in ceramic-glass

在MonteCarlo随机实验方法的基础上,运用基本Metropolis抽样算法动态模拟了微晶玻璃中晶相的生长过程,建立了粒子扩散运移、聚集结晶和脱附三个随机过程的模拟规则及其概率对结晶过程的影响。与以往晶体生长的DLA模拟所不同的是,体系中所有粒子的演化过程是同时进行的,且考虑了体系中微粒间的相互作用势,以它作为结晶和脱附事件发生的概率控制条件,从而得到了与微晶玻璃中实际晶相结构非常近似的模拟结果。

Using the computer simulation technique and Monte Carlo method, under structured FORTRAN language programming environment, crystalline growing in ceramic-glass phase was animated dynamically on the screen in a lattice with 300×200 nodes. The Metropolis sampling technique was chosen to promote the calculating efficiency. The result is consistent with true crystal structure in a kind of CaO-Al2O3-SiO2 system ceramic-glass. Unlike the traditional DLA methods, the grains are simultaneously introduced into the system. A simplified model of interactional potential of grains was founded as a control limit to the diffusion, congregation and desorption probilities.