peterdsharpe/NeuralFoil

NeuralFoil is a practical airfoil aerodynamics analysis tool using physics-informed machine learning, exposed to end-users in pure Python/NumPy.

69
/ 100
Established

Trained on tens of millions of XFoil simulations, NeuralFoil provides 8 model variants trading accuracy for speed, with vectorized evaluation across angle-of-attack and Reynolds number. The hybrid architecture embeds physics-based invariants and includes an `analysis_confidence` metric to flag out-of-distribution predictions, enabling robust design optimization. It integrates seamlessly with AeroSandbox for extended capabilities (compressibility, post-stall behavior, control surfaces) while remaining a standalone NumPy runtime with <500 lines of user-facing code.

383 stars and 11,336 monthly downloads. Available on PyPI.

Maintenance 6 / 25
Adoption 19 / 25
Maturity 25 / 25
Community 19 / 25

How are scores calculated?

Stars

383

Forks

51

Language

Python

License

MIT

Last pushed

Oct 17, 2025

Monthly downloads

11,336

Commits (30d)

0

Dependencies

2

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/peterdsharpe/NeuralFoil"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.