Keel Section

AI-Driven Design for Performance

This case study explores the optimization of 2D hydrofoil sections using machine-learning tools, with application to keel design for high-performance sailing.

Context

Designing foil sections involves trade-offs between hydrodynamic efficiency, structural requirements, and cavitation boundaries. Standard airfoils often don’t meet all constraints, so optimization is required to find the best solution that meets all the requirements.

Methodology

• Input Parameters
  Defined performance and cavitation conditions, along with structural and geometrical constraints (e.g. max chord, thickness).
  For the particular case study:
  • Upwind: 8 kn, CL·chord = 0.16
  • Downwind: 12 kn, CL·chord = 0.05
• Candidate Generation
  A parametric model was used to produce thousands of candidate section shapes.
  Approximately 6,000 candidates were analyzed for the particular study.
• Evaluation & Optimization
  Each section is evaluated using XFoil, and an AI-based surrogate model was trained to speed up the search for an optimal design meeting all constraints.

Results

During the case study, three baseline profiles were considered:

• Eppler
• NACA-6 series
• NACA-0018

The resulting design achieved a ~10% drag reduction compared to the Eppler section, while still satisfying structural constraints.

Summary

Using a combination of parametric modeling, XFoil, and AI-driven optimization, we developed a foil section with improved hydrodynamic performance while complying with structural criteria — a workflow applicable to a broad range of hydro/aerodynamic challenges.