Predicting Time-Dependent Flow Over Complex Geometries Using Operator Networks
physics.flu-dyn, cs.LG
Submitted: 2025-12-04
Updated: 2025-12-04
Journal ref: Computer Methods in Applied Mechanics and Engineering 456 (2026) 118931
DOI: 10.1016/j.cma.2026.118931
Code: https://github.com/baskargroup/TimeDependent-DeepONet
License: http://creativecommons.org/licenses/by/4.0/
The gist: Fast, geometry-generalizing surrogates for unsteady flow remain challenging.
Terminology
Abstract
Fast, geometry-generalizing surrogates for unsteady flow remain challenging. We present a time-dependent, geometry-aware Deep Operator Network that predicts velocity fields for moderate-Re flows around parametric and non-parametric shapes. The model encodes geometry via a signed distance field (SDF) trunk and flow history via a CNN branch, trained on 841 high-fidelity simulations. On held-out shapes, it attains about 5% relative L2 single-step error and up to 1000X speedups over CFD. We provide physics-centric rollout diagnostics, including phase error at probes and divergence norms, to quantify long-horizon fidelity. These reveal accurate near-term transients but error accumulation in fine-scale wakes, most pronounced for sharp-cornered geometries. We analyze failure modes and outline practical mitigations. Code, splits, and scripts are openly released at: https://github.com/baskargroup/TimeDependent-DeepONet to support reproducibility and benchmarking.
Sources
- 3D Neural Operator-Based Flow Surrogates around 3D geometries: Signed Distance Functions and Derivative Constraints
- FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators
- Direct Flow Simulations with Implicit Neural Representation of Complex Geometry
- Temporal Neural Operator for Modeling Time-Dependent Physical Phenomena
- Simulating incompressible flows over complex geometries using the shifted boundary method with incomplete adaptive octree meshes