Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

OptiWing3D: A Diverse Dataset of Optimized Wing Designs

Cashen Diniz, Mark D. Fuge

TL;DR

OptiWing3D provides the first public, high-fidelity 3D wing optimization dataset paired with 2D counterparts, enabling direct 2D–3D comparisons and multi-fidelity studies in data-driven aerodynamic design. The dataset comprises 1552 simulations (776 optimized 3D wings from 1400 extruded airfoils) generated under subsonic to transonic conditions, with detailed geometric, pressure, and flow-condition metadata. A conditional latent diffusion framework with a Bezier autoencoder baseline demonstrates feasible generation of optimized wings from flow constraints, achieving low shape MSE and capturing substantial design diversity, while identifying data-efficiency limits around 350 samples. The work highlights that 3D optimization introduces greater geometric and performance complexity, especially toward the wingtip due to 3D effects, and lays groundwork for future multi-fidelity and inverse-design research.

Abstract

OptiWing3D is the first publicly available dataset of high-fidelity shape optimized 3D wing geometries. Existing aerodynamics datasets are either limited to 2D simulations, lack optimization, or derive diversity solely from perturbations to a single baseline design, constraining their application as benchmarks to inverse design approaches and in the study of design diversity. The OptiWing3D dataset addresses these gaps, consisting of 1552 simulations resulting in 776 wing designs initialized from distinct extruded airfoil cross-sections. Additionally, a majority of the optimized wings in the dataset are paired to 2D counterparts optimized under identical conditions, creating the first multi-fidelity aerodynamic shape optimization dataset. Moreover, this structure allows for a direct comparison between 2D and 3D aerodynamic simulations. It is observed that 3D optimized designs diverge most prominently from the 2D-optimized designs near the wingtip, where three-dimensional effects are strongest, a finding made possible by the paired nature of the dataset. Finally, we demonstrate a constraint-aware conditional latent diffusion model capable of generating optimized wings from flow conditions, establishing a baseline for future inverse design approaches. The dataset, containing wing geometries and surface pressure distributions is publicly released to advance research in data-driven aerodynamic design.

OptiWing3D: A Diverse Dataset of Optimized Wing Designs

TL;DR

OptiWing3D provides the first public, high-fidelity 3D wing optimization dataset paired with 2D counterparts, enabling direct 2D–3D comparisons and multi-fidelity studies in data-driven aerodynamic design. The dataset comprises 1552 simulations (776 optimized 3D wings from 1400 extruded airfoils) generated under subsonic to transonic conditions, with detailed geometric, pressure, and flow-condition metadata. A conditional latent diffusion framework with a Bezier autoencoder baseline demonstrates feasible generation of optimized wings from flow constraints, achieving low shape MSE and capturing substantial design diversity, while identifying data-efficiency limits around 350 samples. The work highlights that 3D optimization introduces greater geometric and performance complexity, especially toward the wingtip due to 3D effects, and lays groundwork for future multi-fidelity and inverse-design research.

Abstract

OptiWing3D is the first publicly available dataset of high-fidelity shape optimized 3D wing geometries. Existing aerodynamics datasets are either limited to 2D simulations, lack optimization, or derive diversity solely from perturbations to a single baseline design, constraining their application as benchmarks to inverse design approaches and in the study of design diversity. The OptiWing3D dataset addresses these gaps, consisting of 1552 simulations resulting in 776 wing designs initialized from distinct extruded airfoil cross-sections. Additionally, a majority of the optimized wings in the dataset are paired to 2D counterparts optimized under identical conditions, creating the first multi-fidelity aerodynamic shape optimization dataset. Moreover, this structure allows for a direct comparison between 2D and 3D aerodynamic simulations. It is observed that 3D optimized designs diverge most prominently from the 2D-optimized designs near the wingtip, where three-dimensional effects are strongest, a finding made possible by the paired nature of the dataset. Finally, we demonstrate a constraint-aware conditional latent diffusion model capable of generating optimized wings from flow conditions, establishing a baseline for future inverse design approaches. The dataset, containing wing geometries and surface pressure distributions is publicly released to advance research in data-driven aerodynamic design.

Paper Structure

This paper contains 29 sections, 11 equations, 18 figures, 4 tables.

Table of Contents

  1. Introduction and Related Work
  2. Dataset Generation
  3. Dataset Summary and Structure
  4. Design Conditions
  5. Wing Parameterization
  6. Mesh Generation and Deformation
  7. Optimization Problem
  8. Solver
  9. Dataset Results
  10. Optimized Designs
  11. 2D-3D Optimization Comparison
  12. Dimensional Analysis
  13. Generative Model Setup
  14. Overview
  15. Diffusion Model Background
  16. ...and 14 more sections

Figures (18)

  • Figure 1: 2D Airfoil and 3D Wing Data Generation
  • Figure 2: Selected Initial (Non-Optimized) Wing Designs From the OptiWing3D Dataset
  • Figure 3: Sample Mesh and FFD Cage
  • Figure 4: Random samples: optimized, initial geometries and surface pressures
  • Figure 5: Aggregate optimized shape and pressure differences
  • ...and 13 more figures