遺伝的アルゴリズムを用いた薄膜の安定構造解析 (<小特集>マイクロスケールの流体・熱流動現象)

概要

論文の詳細を見る
In this paper, molecular simulations using a genetic algorithm are conducted to investigate the stability of deposited atoms on a substrate. First, to confirm the validity of the method, the stability of structures of L-J clusters is studied and it is successfully shown that a stable 13-atoms cluster has a well-known icosahedral structure. Then using this method, the stability of structures of deposited atoms on a solid surface is investigated. The results show that deposited atoms have either an island or a layer structure, depending on the combination of intermolecular potentials of the deposited atom and the substrate atom. Furthermore, to apply the method for systems having a certain temperature, a new method using a modified intermolecular potential is proposed. The results of the advanced simulation agree qualitatively with those of the MD simulation. Although the GA simulation gives good results, it also requires much less CPU time than MD simulation. Therefore, it can be concluded that the GA simulation is a superior technique for examining stable structures of many-particle systems.
一般社団法人日本機械学会の論文
1999-02-25