ChatCFD: An LLM-Driven Agent for End-to-End CFD Automation with Structured Knowledge and Reasoning

TL;DR

ChatCFD is an innovative LLM-based multi-agent system that automates CFD workflows end-to-end, integrating structured knowledge and reasoning to significantly improve success rates, fidelity, and efficiency over existing methods.

Contribution

This work introduces ChatCFD, the first LLM-driven multi-agent system for comprehensive CFD automation with structured knowledge, error localization, and iterative reasoning capabilities.

Findings

Achieves 82.1% success rate on benchmark CFD cases.

Attains 68.12% physical fidelity, surpassing prior methods.

Demonstrates high flexibility in solver selection and complex case reproduction.

Abstract

Computational Fluid Dynamics (CFD) is critical for scientific advancement but is hindered by operational complexity and high expertise barriers. This paper introduces ChatCFD, a Large Language Model (LLM)-driven multi-agent system designed for end-to-end CFD automation using OpenFOAM. Powered by DeepSeek-R1/V3, ChatCFD integrates structured domain knowledge bases, a precise error locator, and iterative reflection to dramatically outperform existing methods. On 315 benchmark cases, ChatCFD achieves 82.1% execution success (vs. 6.2% for MetaOpenFOAM and 42.3% for Foam-Agent) and 68.12% physical fidelity - a novel metric assessing scientific meaningfulness beyond mere runnability. A dedicated Physics Interpreter attains 97.4% summary fidelity, bridging the gap between narrative fluency and the enforcement of tight physical constraints. Resource analysis confirms efficiency, averaging 192.1k tokens and $0.208 per case, significantly lower than baseline costs. Ablation studies identify the Error Locator and Solver Template DB as critical, with the latter's removal collapsing accuracy to 48%. The system exhibits robust flexibility, achieving 95.23% success in autonomous solver selection and 100% in turbulence modeling, while successfully reproducing complex literature cases (e.g., NACA0012, supersonic nozzle) with 60-80% success rates where baselines failed. Featuring a modular, MCP-compatible design, ChatCFD facilitates scalable, collaborative AI-driven CFD. Code is available at: https://github.com/ConMoo/ChatCFD