Research map / Probabilistic and scientific modeling / Molecules and proteins
**TandemFoilSet**: Datasets for Flow Field Prediction of Tandem-Airfoil Through the Reuse of Single Airfoils
Wei Xian Lim, Loh Sher En Jessica, Zenong Li, Thant Zin Oo, Wai Lee Chan, Adams Wai-Kin Kong
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add pre-training on single-airfoil flow fields only (PRE) (ablation) | MGN · TandemFoilSet Cruise AOA=0/5, Takeoff | MSE: 1.03 -> 1.04, 1.34 -> 1.21, 3.74 -> 3.69 | No clear effect |
| add pre-training only (PRE) (ablation) | IVE · TandemFoilSet Cruise AOA=0/5, Takeoff | MSE: 0.85 -> 0.98, 2.53 -> 2.73 | Hurt |
| add smooth-combined fields as input after pre-training (PRE-FREE + COMB) (ablation) | MGN · TandemFoilSet Takeoff | MSE: 3.74 -> 1.31 | Improved |
| add combined-field residual training (RES-FREE + RES-COMB) (ablation) | MGN · TandemFoilSet Cruise AOA=5 | MSE: 1.34 -> 0.68 | Improved |
| add pre-training on top of residual training (PRE-RES-FREE + RES-COMB) (ablation) | MGN · TandemFoilSet Cruise AOA=0 | MSE: 0.49 -> 0.45 | No clear effect |
| add the full curriculum (pre-training, freestream residual, combined residual) (ablation) | IVE · TandemFoilSet Cruise AOA=0/5, Takeoff | MSE: 0.85 -> 0.52, 1.05 -> 0.63, 2.53 -> 0.83 | Improved |
| add directional integrated distance (DID) features to shortest vector (ablation) | MGN · TandemFoilSet Cruise AOA=0 | MSE: 11.51 -> 1.03 | Improved |
| add DID features to shortest vector (ablation) | MGN · TandemFoilSet Takeoff | MSE: 8.17 -> 3.74 | Improved |
| use multi-NN sub-domain inference instead of a single NN (ablation) | MGN + RES-FREE + RES-COMB · TandemFoilSet Cruise AOA=0 | MSE: 1.66 -> 0.49 | Improved |
| use full curriculum instead of baseline under Re extrapolation | MGN · TandemFoilSet Cruise Random (Re extrapolation) | MSE: 4.85 -> 0.36 | Improved |
| use curriculum instead of multi-task joint training on single and tandem data (ablation) | MGN · TandemFoilSet Cruise Random | MSE: 1.80 -> 0.10 | Improved |
| use full curriculum on Race Car data | MGN · TandemFoilSet Race Car (uniform split) | MSE: 0.61 -> 0.21 | Improved |
Closest papers by meaning
- Aerodynamic Coefficients Prediction via Cross-Attention Fusion and Physical-Informed Training (AAAI 2025 · similarity 0.86)
- AFBench: A Large-scale Benchmark for Airfoil Design (NeurIPS 2024 · similarity 0.85)
- Deep Learning-based Algebraic Reynolds Stress Closures for RANS simulations of Turbulent Flows (NeurIPS 2026 · similarity 0.83)
- FuncGenFoil: Airfoil Generation and Editing Model in Function Space (NeurIPS 2025 · similarity 0.82)
- DrivAerNet++: A Large-Scale Multimodal Car Dataset with Computational Fluid Dynamics Simulations and Deep Learning Benchmarks (NeurIPS 2024 · similarity 0.82)
- WindsorML: High-Fidelity Computational Fluid Dynamics Dataset For Automotive Aerodynamics (NeurIPS 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
