WangZhuo
Chengdu CaiZhi Technology Co., Ltd, Chengdu, 610041, China 

EXTENDED ABSTRACT: Materials science and engineering are key components of the modem industrial field, and they have a direct impact on product performance, sustainability and innovation. With the development of science and technology, materials science has entered the era of materials digital intelligence using new technologies such as data, computing and AI to accelerate the design and optimisation of materials and realise the intelligence of materials science research. This paper describes the practice of materials digital intelligence in industrial applications from three aspects of data, computing and AI applications.
In terms of data, with the continuous upgrading of data transmission and storage technology and the development of high-throughput preparation and characterisation technology, rapid screening of new materials and rapid accumulation of material data can be realised, so as to optimise material components and processes and shorten the product development cycle. For example, by using data acquisition and cleaning technology, the R&D related data such as inspection data, product standards, quality objections, user files, quality points, technical agreements, metallurgical specifications, etc. dispersed in the inspection system, production, supply and marketing system of Daye Special Steel can be collected and cleaned, which is convenient for R&D personnel to query and make use of the data, and on the basis of which, the user file management system and R&D project management system are constructed, which forms a user-oriented variety development work and R&D project management system. User-oriented variety development work and R&D project management.
On the computational side, tools such as first principles, molecular dynamics, thermodynamics and finite element analysis have played a key role in materials design. Meanwhile, with the development of high-throughput computing technology, multi-scale computational material science models can be established from the atomic scale to the macroscopic scale, and the use of these computational models can achieve the global search and optimal design of material components and structures. For example, the multi-scale materials integrated design and computation platform established in the Digital R&D Centre of China Steel Research Group has realized the integrated scheduling and high-throughput computation application ofVASP, Thermo-Cale, JMatPro and other computational software, which accelerates the design process of materials application and enhances the efficient reuse of computational data accumulation and computational knowledge resource sharing services.
Most importantly, AI technology has made its mark in materials research. With the development of optimised machine learning algorithms, they can be used to rapidly analyse and interpret complex materials performance data and discover hidden correlations in the data. Optimisation of the training of machine learning models and the incorporation of active learning can be used to predict the properties of new materials and optimise the design of material components and processes. For example, machine learning is used to predict the effects of thermal deformation process and heat treatment process on the organisation and properties of high temperature alloys, to construct an integrated design scheme for the composition and process of high strength and high toughness high temperature alloys for small reactors with tubes suitable for reactor applications, and to accelerate the development rate of high temperature alloys for small reactors.
The digital intelligence of materials based on data, computing and AI promotes the rapid development of the new materials industry by transforming the materials "trial and error" research and development mode, and its industrial application is rapidly changing the face of materials science and engineering. This trend brings greater speed, efficiency and innovation to materials research and development, and is expected to realise performance breakthroughs, shorter development cycles and lower development costs in various material fields. 

Keywords: data; Integrated Computational Materials Science; AI; materials digitalisation; applied practice;