材料信息学及其在材料研究中的应用

Materials Informatics and Its Application in Materials Research

2011年美国奥巴马总统提出的材料基因组计划(MGI),旨在以比原先至少快两倍的速度开发和制造先进材料,且成本仅为原先的几分之一,这促使了材料信息学的快速发展。材料信息学是信息学技术在材料学中的应用,通过建设材料信息数据库、集成材料研究设计平台和材料数据挖掘方法对材料大数据进行分析和预测,快速发现决定材料性能的"基因",也就是材料成分-工艺-组织-性能之间的定量关系,可以有效地加快材料研发设计。介绍了材料信息学的基本概念和主要研究领域,描述了材料信息学中的3个主要组成部分:材料信息数据库、集成材料设计平台和材料数据挖掘技术的主要内容和应用实例。材料信息数据库储存和管理各类材料数据,包括材料基础性能、晶体结构数据、模拟计算数据、试验与工艺数据、专利数据和各类出版物等;集成材料设计平台提供各种模拟计算方法,如第一性原理、分子动力学、CALPHAD方法、相场模拟和有限元分析;数据挖掘是统计学、机器学习、信息学、可视化技术等学科的交叉领域,是从大数据中发现知识的实用方法。并介绍了成都材智科技搭建的"材智云"集成材料设计平台的框架和功能。思考了材料信息学在材料领域中应用时所面临的难题。

The Materials Genome Initiative （MGI） project proposed by President Obama in 2011 is aimed at two times faster of developing and ,nannfacturing of advanced materials while at a fraction of cost than before. This project promoted the rapid development of materials informatics that is the application of information technology in materials science. Material information database, integrated materials design platform and data milfing methods are used to analyze and predict materials big data, and further reveal the quantitative relationships of the constituents-process-microstmcture-properties in materials science, which is the gene to determine the materials properties. The design and development of advanced materials can be effectively speeded up by materials informaties. This article describes the concepts and main research areas of materials informatics. The materials information databases provide the storage and nmnagement sen,ice of materials data, such as crystal structure data, simulated and predicted da- ta, experimental and processing data, and even patent data and various kinds of publications. Integrated materials design platform provides multiscale simulation techniques of materials research, such as the first- principles calculation, molecular dynamics, CALPHAD method, phase-field simulation and finite element a- nalysis, etc. The calculated data can be added to the materials databases. Data Mining is an interdisciplinary field merging methods from statistics, machine learning, information science, visualization and other disciplines. It is a very useful approach to discover knowledge from materials big data produced by combination of experiments and high throughput calculation. The application of nmterials databases, integrated materials design platform and data mining in malerials research are introduced. Especially the framework and function of ＂MATGENE＂ integrated materials design platform built by Matclouds technoloo＂ in Chengdu are also described. Finally the challenges of materials intormatics in n laterials research are discussed.