多くの接着継手の破壊力学パラメータは有限要素法により求められている。接着剤が弾性体として取り扱える場合は破壊力学パラメータとしてエネルギー解放率が使われことが多い。しかし，最近の高いじん性化された接着剤では接着層の応力―ひずみ曲線の非線形性を考慮する必要がある。そのような場合には，破壊力学パラメータとしてJ 積分が使用される。本稿ではJ 積分についての概要説明を行った後，有限要素法によるJ 積分の解析法について，具体例を示して解説した。

接着のような界面現象を分子シミュレーションで取り扱うための各種手法について検討する。はじめに，表面をモデル化する際の本質的な困難について議論する。次に，全原子分子動力学法について，パラメータ設定の仕方やソフトウェアを含めて示す。最近，利用の増えてきた化学反応を含む分子動力学法は接着現象を取り扱うためには重要である。高分子系を取り扱うための粗視化分子シミュレーションについても述べる。 最後に，ミクロンスケールの系において表面粗さ，熱輸送や相転移を伴う現象を取り扱うためのシミュレーション手法を示す。

In recent years, the importance of numerical analysis of adhesively bonded parts is increasing with increasing demand for adhesive joints. Cohesive zone model（ CZM） is one of the promising methods to evaluate the crack propagation and strength prediction of the adhesive joints. Using the cohesive element in the finite element method（ FEM）, the fracture process of the adhesive layer can be numerically analyzed. However, a clear understanding of CZM parameters and the input of the appropriate values are essential to creating correct numerical models. In this paper, a traction-separation law was explained based on its physical mechanism. Then, the meaning of the CZM parameters and determination processes of them were explained. Additionally, recent researches of adhesively bonded joints using CZM analysis were reviewed.

General introduction to molecular orbital（ MO） method is afforded with emphasis of the actual utilizations. In this respect, the density functional theory（ DFT） method is mainly employed here in particular among other MO calculation methods such as Hartree-Fock（ HF） and the post HF ones. The way of preparation of the environment for the MO calculation, actually available softwares, and the way of inputs are to be first explained. Several examples revealing the utility of MO calculations are also mentioned: these include achievement of the energetically optimized molecular structure, its geometrical parameters, vibration analyses including the IR and the Raman scattering absorption intensities, thermochemical analyses including the temperature corrections, analyses of the MO patterns, the total electron density, atomic net charge, bond order, and spin densities. Furthermore, estimation of the NMR chemical shift and description of the singlet excitation spectrum are also afforded. Especially in the last part of the article, the theme concerning the adhesion is mentioned and importance of the long-range interaction between molecules as well as the calculation method for which is also to be described.

We introduced the Self-Organizing Map（ SOM） as a visualization tool utilized for a multi-objective material optimization. The SOM, one of un-supervised machine learning techniques, enables to locate and cluster materials on a low-dimensional space according to similarity of multi-objective functions. In this paper, we showed the effectiveness of the SOM through applications to structural materials such as composite and thermosetting resins.