Multiscale modeling of materials and neural operators

TL;DR

This paper introduces neural operators as a powerful tool for multiscale modeling of materials, demonstrating their effectiveness through three illustrative examples.

Contribution

It presents the application of neural operators to multiscale material modeling, highlighting their potential to transfer information across scales.

Findings

Neural operators are effective in multiscale modeling.

Three examples demonstrate neural operators' capabilities.

Neural operators offer discretization-independent modeling solutions.

Abstract

Multiscale modeling is essential for understanding the complex behavior of materials. However, accurately transferring all relevant information from one scale to another has remained an outstanding challenge. Neural operators, discretization-independent generalizations of neural networks, is proving to be a powerful tool in addressing this challenge. This article provides an introduction to neural operators, and illustrates their use in multiscale modeling of materials through three selected examples.