Rich multi-dimensional correlative imaging

Abstract

Abstract The use of microstructural design to tailor materials properties has increased sharply in recent years. In parallel the number and the capability of techniques able to characterise materials microstructures has increased sharply too providing structural, chemical and crystallographic information. Here we examine how correlated 3D, 4D (3D + time) and multi-dimensional imaging enable a much richer picture to be built up of a materials microstructure. We look at how a data-centric approach can support the use of materials informatics, digital twinning and machine learning to accelerate the design of new materials systems and to optimise the manufacturing of established ones. However for this to happen we need to develop ways to digitally fingerprint the microstructural images and maps we collect such that they can be incorporated into machine learning schemes. Through the use of case studies (multimodal imaging) we look at correlative imaging across scales, across time (the dilation of electrode materials in lithium batteries during discharging and fast corrosion of magnesium), as well as across multiple modalities (butterfly defects in bearings steels and the sintering and recrystallization of powders). These demonstrate how different techniques can come together to provide complementary aspects of the bigger picture.