Show HN: Autonomous CAD design and OpenFOAM optimization loop using local LLMs

Category: ai-ml

Tags: multi-agent, llm, cad, cfd, optimization, openfoam, parametric-design

Score: 6.3/10 (Innovation: 6, Technical: 8, Documentation: 7, Utility: 4)

This project implements a multi-agent system that uses local LLMs (via Ollama) to autonomously propose, simulate, and optimize 3D-printable quadcopter propellers through CAD (CadQuery) and CFD (OpenFOAM). It's interesting because it demonstrates a practical, trustworthy pattern of letting AI suggest designs while using deterministic physics code for scoring, avoiding common pitfalls of LLM hallucination.

Target audience: backend devs, data engineers, mechanical engineers

Repository: https://github.com/ostenjap/LLM-Agent-generated-Quadcopter-Prop · Python · MIT

View on Hacker News