MultiLevel variational MultiScale (ML-VMS) framework for large-scale simulation

Lei Zhang

doi:10.1016/j.cma.2026.118807

MultiLevel variational MultiScale (ML-VMS) framework for large-scale simulation

2026-02-13

0

PRE

AI

DOI：10.1016/j.cma.2026.118807

Original

Original request for help

Save

Abstract

En 中文

• A novel multiscale framework that resolves disparate spatial and temporal scales using linear finite element meshes. • An efficient approach with arbitrary interpolation orders by adjusting hyperparameters of Convolution Hierarchical Deep-learning Neural Network without increasing DoFs. • A space-time formulation provides unconditional stability, solving all time steps at once.

Keywords:

Multiscale framework

Space-time formulation

Variational methods

Deep-learning neural network

Large-scale simulation

MultiLevel variational MultiScale (ML-VMS) framework for large-scale simulation

Abstract

Journal

Organization