CFDTwin v0.2.0
This release packages CFDTwin as an open-source Python toolkit for building neural-network surrogate models from ANSYS Fluent simulations, with the desktop GUI available as an optional install extra.
Installation
API-only installation:
pip install cfdtwinAPI + desktop GUI:
pip install "cfdtwin[gui]"The API-only install avoids the Qt/PySide6 GUI dependency and is better suited for headless, HPC, and scripted workflows. The GUI extra installs the wizard-based desktop app for interactive Setup -> DOE -> Simulate -> Train -> Analyze workflows.
Highlights
- Separates the lightweight API install from the optional GUI dependency.
- Provides the same CFDTwin workflow through both Python scripts and the desktop GUI.
- Supports Fluent case setup, input/output definition, DOE generation, batch simulation, surrogate training, validation, and prediction.
- Includes documentation and a four-part GUI walkthrough video series.
- Supports scalar, surface-field, and cell-zone output workflows for CFD surrogate modeling.
Documentation
- Documentation: https://uark-ned3.github.io/CFDTwin/
- Repository: https://github.com/UARK-NED3/CFDTwin
- License: MIT
Citation
This release is archived on Zenodo:
If you use CFDTwin in research, please cite this archived release and the accompanying software paper/preprint when available.