High-Throughput Computing Assisted by Knowledge Graph to Study the Correlation between Microstructure and Mechanical Properties of 6XXX Aluminum Alloy.

Abstract

It is of great academic and engineering application to study the evolution of microstructure and properties of age-strengthened aluminum alloys during heat treatment and to establish quantitative prediction models that can be applied to industrial production. The main factors affecting the peak aging state strength of age-strengthened aluminum alloys are the precipitates, solid solution elements, grain size effects, and textures formed during the material processing. In this work, these multi-scale factors are integrated into the framework of the knowledge graph to assist the following crystal plasticity finite elements simulations. The constructed knowledge graph is divided into two parts: static data and dynamic data. Static data contains the basic properties of the material and the most basic property parameters. Dynamic data is designed to improve awareness of static data. High-throughput computing is performed to further obtain clear microstructure-property relationships by varying the parameters of materials properties and the characteristics of the structure models. The constructed knowledge graph can be used to guide material design for 6XXX Al-Mg-Si based alloys. The past experimental values are used to calibrate the phenomenological parameters and test the reliability of the analysis process.