Welcome to DeepFlame’s Documentation

DeepFlame is a deep learning empowered computational fluid dynamics package for single or multiphase, laminar or turbulent, reacting flows at all speeds. It aims to provide an open-source platform to combine the individual strengths of OpenFOAM, Cantera and PyTorch libraries for deep learning assisted reacting flow simulations. It is also has the scope to incorporate next-generation heterogenous supercomputing and AI acceleration infrustructures such as GPU and FPGA.

The deep learning algorithms and models used in the DeepFlame tutorial examples are developed and trained independently by our collaborators team – DeepCombustion. Please refer to their website for detailed information.

Note

This project is under active development.

Quick Start

• 1. Installation
  • 1.1. Prerequisites
  • 1.2. Configure
  • 1.3. Build and Install
  • 1.4. Other Options
• 2. Two Examples
  • 2.1. DeepFlame with DNN
  • 2.2. DeepFlame without DNN
• 3. Brief Introduction to Inputs

Solvers

• 1. df0DFoam
  • 1.1. Zero-Dimensional ignition reactor
• 2. dfLowMachFoam
  • 2.1. One-Dimensional Planar Flame
  • 2.2. Two-Dimensional Jet Flame
  • 2.3. Three-Dimensional reactive Taylor-Green Vortex
• 3. dfHighSpeedFoam
  • 3.1. One-Dimensional Reactive Shock Tube
  • 3.2. One-Dimensional H2/Air Detonation
  • 3.3. Two-Dimensional H2/Air Detonation
• 4. dfSprayFoam
  • 4.1. aachenBomb

Utility

• 1. Reaction Mechanism Convertion
• 2. Flame Speed

Project Details

• 1. Developers Team
• 2. Collaborators Team
• 3. How to Cite
• 4. License

Contributing to DeepFlame

• Submitting a Pull Request