I-FENN with DeepONets: Accelerating simulations in coupled multiphysics problems

Fouad Amin

doi:10.1016/j.cma.2025.118645

I-FENN with DeepONets: Accelerating simulations in coupled multiphysics problems

2025-12-23

0

PRE

AI

DOI：10.1016/j.cma.2025.118645

Original

Original request for help

Save

Abstract

En 中文

• We introduce a GRU-DeepONet-based I-FENN framework for multiphysics simulations, demonstrated in thermoelasticity and poroelasticity. • A modified MIONet accommodates multiple input functions, with a new strategy for enforcing boundary conditions on distinct segments. • Three case studies validate the framework’s generalization to unseen loading scenarios. • The approach is applicable in 2D and 3D, mesh-independent and scalable, as confirmed by mesh-refinement tests. • Implementation leverages the deal.II library for efficient HPC deployment with optimized data transfer and memory use.

I-FENN with DeepONets: Accelerating simulations in coupled multiphysics problems

Abstract

Journal

Organization