CFD Forge: A Physics-Constrained Agentic Framework for Autonomous CFD
Abstract
Setting up, running, and assessing a computational fluid dynamics (CFD) simulation requires substantial manual effort, from meshing and solver setup to diagnosis and interpretation, and repeating it across design variants slows engineering iteration. We present CFD Forge, an agentic framework built around OpenFOAM that carries an engineering request through case construction, meshing, solver execution, diagnostics, and reporting. A large language model interprets the request, reads the simulation evidence, identifies likely causes of problems, and proposes actions from a closed set. Automation is only useful if its results can be trusted, so acceptance is not left to the model: deterministic checks, registered for each type of flow, measure explicit criteria and decide. We demonstrate the framework on a converging-diverging nozzle and a two-dimensional forward-facing step, including geometry and operating changes requested in text. A turbulent cube, run outside the agent loop, shows why the checks matter: the solver completed and the mean drag settled, but the lateral-force oscillations were still growing, so a retrospective assessment withheld acceptance. To test this separation, we replayed 36 recorded simulation states, some with deliberately altered diagnostics, through the same model in two setups. Judging from the evidence without the check results, the model accepted 12 of the 13 detectable faults in all three repeats; with CFD Forge's checks deciding, none was accepted. The two setups also differ in prompting, and a fixed rule reached the same verdicts, so the model's value lies in the open-ended work rather than the verdict. Case files, recorded model calls, and decisions are retained for audit. A model that does the work, paired with checks that decide what is accepted, is a step toward automated CFD that can be trusted in high-throughput design and active-learning loops.
Related
- From Automated Simulation to Autonomous Discovery: A Hierarchical Framework for Agentic Computational Materials Science
- HEPTABench: A framework for portable and controlled agentic evaluations in high energy physics
- Solver Agent: an Agentic AI Framework for Theoretical Physics Computations Applied to F-theory Uplifts of O3-planes and S-folds
- Autonomous Quantum Measurement with Agentic AI
- Autonomous Analysis of Thermoelectric Materials Big Data by Multi-Agentic AI and Construction of a Knowledge Base and Specialized AI
BibTeX
@misc{oza2026forge,
title = {{CFD Forge: A Physics-Constrained Agentic Framework for Autonomous CFD}},
author = {Aakash Shailesh Oza and Ryan F. Johnson and Amir Barati Farimani},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.14104},
eprint = {2610.14104},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.14104},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - CFD Forge: A Physics-Constrained Agentic Framework for Autonomous CFD
AU - Aakash Shailesh Oza
AU - Ryan F. Johnson
AU - Amir Barati Farimani
PY - 2026
DA - 2026/10/08
PB - arXiv
DO - 10.48550/arxiv.2610.14104
UR - https://arxiv.org/abs/2610.14104
AB - Setting up, running, and assessing a computational fluid dynamics (CFD) simulation requires substantial manual effort, from meshing and solver setup to diagnosis and interpretation, and repeating it across design variants slows engineering iteration. We present CFD Forge, an agentic framework built around OpenFOAM that carries an engineering request through case construction, meshing, solver execution, diagnostics, and reporting. A large language model interprets the request, reads the simulation evidence, identifies likely causes of problems, and proposes actions from a closed set. Automation is only useful if its results can be trusted, so acceptance is not left to the model: deterministic checks, registered for each type of flow, measure explicit criteria and decide. We demonstrate the framework on a converging-diverging nozzle and a two-dimensional forward-facing step, including geometry and operating changes requested in text. A turbulent cube, run outside the agent loop, shows why the checks matter: the solver completed and the mean drag settled, but the lateral-force oscillations were still growing, so a retrospective assessment withheld acceptance. To test this separation, we replayed 36 recorded simulation states, some with deliberately altered diagnostics, through the same model in two setups. Judging from the evidence without the check results, the model accepted 12 of the 13 detectable faults in all three repeats; with CFD Forge's checks deciding, none was accepted. The two setups also differ in prompting, and a fixed rule reached the same verdicts, so the model's value lies in the open-ended work rather than the verdict. Case files, recorded model calls, and decisions are retained for audit. A model that does the work, paired with checks that decide what is accepted, is a step toward automated CFD that can be trusted in high-throughput design and active-learning loops.
ER -