Predictive Scale-Bridging Simulations through Active Learning

TL;DR

This paper introduces an active learning framework that enhances scale-bridging simulations by efficiently integrating fine-scale molecular dynamics data into coarse-scale models, improving physical fidelity in computational science applications.

Contribution

It presents a novel active learning method for dynamically coupling fine-scale molecular simulations with coarse-scale hydrodynamics, addressing uncertainty quantification and trajectory forecasting.

Findings

Improved accuracy in coarse-scale predictions through active learning

Efficient use of molecular dynamics simulations to inform larger-scale models

Enhanced uncertainty quantification in neural network-based simulations

Abstract

Throughout computational science, there is a growing need to utilize the continual improvements in raw computational horsepower to achieve greater physical fidelity through scale-bridging over brute-force increases in the number of mesh elements. For instance, quantitative predictions of transport in nanoporous media, critical to hydrocarbon extraction from tight shale formations, are impossible without accounting for molecular-level interactions. Similarly, inertial confinement fusion simulations rely on numerical diffusion to simulate molecular effects such as non-local transport and mixing without truly accounting for molecular interactions. With these two disparate applications in mind, we develop a novel capability which uses an active learning approach to optimize the use of local fine-scale simulations for informing coarse-scale hydrodynamics. Our approach addresses three challenges: forecasting continuum coarse-scale trajectory to speculatively execute new fine-scale molecular dynamics calculations, dynamically updating coarse-scale from fine-scale calculations, and quantifying uncertainty in neural network models.