Interfacing finite elements with deep neural operators for fast multiscale modeling of mechanics problems

Minglang Yin,Enrui Zhang,Yue Yu,George Em Karniadakis

DOI: https://doi.org/10.1016/j.cma.2022.115027

IF: 6.588

2022-05-11

Computer Methods in Applied Mechanics and Engineering

Abstract:Multiscale modeling is an effective approach for investigating multiphysics systems with largely disparate size features, where models with different resolutions or heterogeneous descriptions are coupled together for predicting the system's response. The solver with lower fidelity (coarse) is responsible for simulating domains with homogeneous features, whereas the expensive high-fidelity (fine) model describes microscopic features with refined discretization, often making the overall cost prohibitively high, especially for time-dependent problems. In this work, we explore the idea of multiscale modeling with machine learning and employ DeepONet, a neural operator, as an efficient surrogate of the expensive solver. DeepONet is trained offline using data acquired from the fine solver for learning the underlying and possibly unknown fine-scale dynamics. It is then coupled with standard PDE solvers for predicting the multiscale systems with new boundary/initial conditions in the coupling stage. The proposed framework significantly reduces the computational cost of multiscale simulations since the DeepONet inference cost is negligible, facilitating readily the incorporation of a plurality of interface conditions and coupling schemes. We present various benchmarks to assess the accuracy and efficiency, including static and time-dependent problems. We also demonstrate the feasibility of coupling of a continuum model (finite element methods, FEM) with a neural operator, serving as a surrogate of a particle system (Smoothed Particle Hydrodynamics, SPH), for predicting mechanical responses of anisotropic and hyperelastic materials. What makes this approach unique is that a well-trained over-parametrized DeepONet can generalize well and make predictions at a negligible cost.

mechanics,engineering, multidisciplinary,mathematics, interdisciplinary applications

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to effectively combine the low - precision large - scale model and the high - precision small - scale model in multi - scale mechanical system modeling to reduce the computational cost while maintaining the prediction accuracy. Specifically, the paper explores the method of using Deep Neural Operator (DeepONet) as an effective substitute for expensive solvers to achieve fast multi - scale modeling. Traditionally, the fine - scale model in multi - scale modeling requires fine discretization to describe microscopic features, which leads to an excessively high overall computational cost, especially in time - dependent problems. By training DeepONet to learn the fine - scale dynamics and coupling it with the standard Partial Differential Equation (PDE) solver, the computational cost of multi - scale simulation can be significantly reduced, because the inference cost of DeepONet is almost negligible, so that various interface conditions and coupling schemes can be easily combined in the coupling stage.

The main contribution of the paper lies in proposing a new multi - scale coupling framework, which combines deep - learning techniques with traditional numerical methods to predict the mechanical response of materials with complex microstructures. In particular, the paper shows how to couple the Finite Element Method (FEM) with the neural operator, which acts as a proxy for particle systems (such as Smoothed Particle Hydrodynamics, SPH), to predict the mechanical response of anisotropic and hyperelastic materials. The uniqueness of this method is that a fully - trained over - parameterized DeepONet can generalize well and make predictions at an extremely low cost.