Predicting Time-Dependent Flow Over Complex Geometries Using Operator Networks
Organizations: Iowa State University, Ames, Iowa, USA
Abstract
Developing neural surrogates for unsteady flow that generalize across complex geometries and maintain accuracy over extended rollouts remains a critical challenge for accelerating flow simulations. 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, and is trained using 841 high-fidelity simulations from the FlowBench dataset (https://baskargroup.bitbucket.io/). The model attains relative L2 single-step error and up to 1000x speedups over CFD on 262 test geometries not used for training. We provide physics-centric rollout diagnostics, including probe-based phase lag and Strouhal frequency analysis, to quantify long-horizon fidelity. These reveal accurate near-term transients but systematic error accumulation in fine-scale wakes, most pronounced for sharp-cornered geometries. We analyze failure modes and outline practical mitigation strategies including physics-informed regularization and diffusion-based refinement. Code, splits, and scripts are openly released at: https://github.com/baskargroup/TimeDependent-DeepONet to support reproducibility and benchmarking.
Figures & tables
| mean phase lag [rad] | max phase lag [rad] | ||||
|---|---|---|---|---|---|
| 1 | 1 | 0.233 | 0.563 | 0.337 | 3.004 |
| 4 | 0.228 | 0.546 | 0.353 | 2.876 | |
| 8 | 0.210 | 0.521 | 0.128 | 2.886 | |
| 16 | 0.190 | 0.418 | 0.168 | 2.666 | |
| 2 | 1 | 0.209 | 0.589 | 0.267 | 2.807 |
| 4 | 0.210 | 0.575 | 0.365 | 2.832 |
Appendix figures & tables4 assets
Supplementary material from the paper’s appendix.
Appendix
| Component | Configuration |
|---|---|
| Branch CNN Input | |
| CNN Encoder | 3 Parallel conv streams ( , , ), each with max pooling, producing ; concatenated to . |
| Encoder Fusion Stage 1 | conv ( ), pool . |
| Encoder Fusion Stage 2 | conv ( ), pool . |
| Branch MLP (Stage 1) | , ReLU |
| Trunk Input | with channels corresponding to |
| Hyperparameter | |
| Latent dimension | 64 |
| Hidden dimension (MLPs) | 128 |
| MLP depth | 2 layers |
| Message-passing steps | 15 |
| Optimizer | Adam |
| Learning rate |
PDE Surrogate Modeling
PDE surrogate modeling builds fast approximations of partial differential equation solutions for simulation, prediction, and design.
See all 363 papers