Machine learning for structure-guided materials and process design

TL;DR

This work presents a holistic and generic optimization approach that covers the entire process-structure-property chain in materials engineering and employs machine learning to address two critical identification problems: a materials design problem, which involves identifying near-optimal material microstructures that exhibit desired properties, and a process design problem that is to find an optimal processing path to manufacture these microstructures.

Abstract

In recent years, there has been a growing interest in accelerated materials innovation in the context of the process-structure-property chain. In this regard, it is essential to take into account manufacturing processes and tailor materials design approaches to support downstream process design approaches. As a major step into this direction, we present a holistic and generic optimization approach that covers the entire process-structure-property chain in materials engineering. Our approach specifically employs machine learning to address two critical identification problems: a materials design problem, which involves identifying near-optimal material microstructures that exhibit desired properties, and a process design problem that is to find an optimal processing path to manufacture these microstructures. Both identification problems are typically ill-posed, which presents a significant challenge for solution approaches. However, the non-unique nature of these problems offers an important advantage for processing: By having several target microstructures that perform similarly well, processes can be efficiently guided towards manufacturing the best reachable microstructure. The functionality of the approach is demonstrated at manufacturing crystallographic textures with desired properties in a simulated metal forming process.

Figures (10)

• Figure 1: Process-structure-property chain following olson1997computational
• Figure 2: General concept for structure-guided materials and process design. The first step (materials design) is addressed by the SMTLO approach, while the second step (process design) is addressed by the MEG-SGGPO approach.
• Figure 3: The neural networks-based Siamese multi-task learning model (left) and the optimization part (right). Together, both parts build the SMTLO approach to solve materials design probplems.
• Figure 4: Projections of the properties space showing the data distribution and the target region. The distribution of the underlying point cloud is displayed in blue, while the gray dots mark data points that are already located inside the target region. The black isolines indicate regions with the same point cloud density.
• Figure 5: Loss curves of the Siamese multi-task learning model for the test dataset (blue) and the training dataset (red) across tasks: (a) validity, (b) reconstruction, (c) regression, and (d) distance preservation, plotted on a logarithmic scale.