ERI: Learning the Constitutive Equations of Chemo-Mechanics from Atomistic Simulations

Utah State University, Logan UT

Investigators

Abstract

This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2). In this Engineering Research Initiation (ERI) project, the governing laws to describe the mechanics of materials in response to applied force is studied. These laws of mechanics are also known as constitutive relations. Traditionally, the constitutive relations are measured from experiments or based on conservation relations. With the development of materials science, new materials are emerging to achieve multiple functionalities. As a result, the constitutive relations are becoming more complicated. For example, materials need to work under electrical fields or chemical potentials. We need to know how materials respond to different external stimuli. The traditional ways of building constitutive relations would no longer work. This project is to develop a new way to build constitutive relations by learning from simulation data. Here, the constitutive relations for silicon anodes in lithium-ion batteries will be constructed. The study will help us better understand electrode materials and how they fail. The new approach for constitutive modeling can potentially uncover new knowledge in materials. It will be an important step towards the design of high-capacity rechargeable batteries. Therefore, results from this project will contribute to the development of the national economy and sustainability. This project will support undergraduate and graduate students, and particularly encourage underrepresented students to join. The supported students will gain research experience across mechanics, computations, data science, and energies. In this project, the sparse regression method will be used to learn the governing differential equations of chemo-mechanics for the silicon anodes in lithium-ion batteries. Sparse regression can identify the functional forms of the constitutive equations entirely based on spatiotemporal data which are obtained from Monte Carlo molecular dynamics simulations. This research includes three tasks: (i) learning the chemo-mechanics equations for bulk diffusion to compare with existing experiments and models for validation purposes; (ii) learning the chemo-mechanics equations for surface diffusion to gain new understandings of the coupling relations of surface diffusion and mechanics; (iii) learning the probabilistic chemo-mechanics equations to demonstrate the capabilities of the data-driven method in quantifying the uncertainties of material behavior. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.