Abstract:Recent developments in deep learning have brought many inspirations for the scientific computing community and it is perceived as a promising method in accelerating the computationally demanding reacting flow simulations. In this work, we introduce DeepFlame , an open-source C++ platform with the capabilities of utilising machine learning algorithms and pre-trained models to solve for reactive flows. We combine the individual strengths of the computational fluid dynamics library OpenFOAM, machine learning framework Torch, and chemical kinetics program Cantera. The complexity of cross-library function and data interfacing (the core of DeepFlame ) is minimised to achieve a simple and clear workflow for code maintenance, extension and upgrading. As a demonstration, we apply our recent work on deep learning for predicting chemical kinetics (Zhang et al. Combust. Flame vol. 245 pp. 112319, 2022 ) to highlight the potential of machine learning in accelerating reacting flow simulation. A thorough code validation is conducted via a broad range of canonical cases to assess its accuracy and efficiency. The results demonstrate that the convection-diffusion-reaction algorithms implemented in DeepFlame are robust and accurate for both steady-state and transient processes. In addition, a number of methods aiming to further improve the computational efficiency, e.g. dynamic load balancing and adaptive mesh refinement, are explored. Their performances are also evaluated and reported. With the deep learning method implemented in this work, a speed-up of two orders of magnitude is achieved in a simple hydrogen ignition case when performed on a medium-end graphics processing unit (GPU). Further gain in computational efficiency is expected for hydrocarbon and other complex fuels. A similar level of acceleration is obtained on an AI-specific chip – deep computing unit (DCU), highlighting the potential of DeepFlame in leveraging the next-generation computing architecture and hardware. Program summary Program Title: DeepFlame CPC Library link to program files: https://doi.org/10.17632/3pg9xmypp3.1 Developer's repository link: https://github.com/deepmodeling/deepflame-dev Licensing provisions: GPLv3 Programming language: C++ Nature of problem: Solving chemically reacting flows with direct (quasi-direct) simulation methods is usually troubled by the following problems: 1. as the widely-used computational fluid dynamics (CFD) toolbox, OpenFOAM features poor ODE solvers for chemistry and oversimplified transport models, yielding non-negligible errors in simulation results; 2. the chemical source term evaluation is the most computationally expensive and usually accounts for more than 80% of total computing time. Solution method: An open-source platform bringing together the individual strengths of OpenFOAM, Cantera and PyTorch libraries is built in this study. In the present implementation, CVODE solvers, detailed transport models and deep learning algorithms are all adopted to assist the simulation of reacting flow. Note that here machine learning is introduced in combination with heterogeneous computing to accelerate the most demanding solving procedure for chemical source term evaluation.

Learning for CasADi: Data-driven Models in Numerical Optimization

CasADi: a software framework for nonlinear optimization and optimal control

Accelerated Optimization in Deep Learning with a Proportional-Integral-derivative Controller

Learning Controllable Adaptive Simulation for Multi-resolution Physics

DeepFlame : A deep learning empowered open-source platform for reacting flow simulations

TorchDA: A Python package for performing data assimilation with deep learning forward and transformation functions

A Data-Centric Optimization Framework for Machine Learning

DeepFlame: A deep learning empowered open-source platform for reacting flow simulations

Lettuce: PyTorch-based Lattice Boltzmann Framework

In Situ Framework for Coupling Simulation and Machine Learning with Application to CFD

L4acados: Learning-based models for acados, applied to Gaussian process-based predictive control

An Automated Machine Learning-Genetic Algorithm Framework With Active Learning for Design Optimization

TorchMD: A deep learning framework for molecular simulations

Learning a Neural Solver for Parametric PDE to Enhance Physics-Informed Methods

CusADi: A GPU Parallelization Framework for Symbolic Expressions and Optimal Control

Learning to optimize: A tutorial for continuous and mixed-integer optimization

ODE-based Learning to Optimize

lfads-torch: A modular and extensible implementation of latent factor analysis via dynamical systems

Finite Operator Learning: Bridging Neural Operators and Numerical Methods for Efficient Parametric Solution and Optimization of PDEs

Reducing Spatial Discretization Error on Coarse CFD Simulations Using an OpenFOAM-Embedded Deep Learning Framework

Physical Design using Differentiable Learned Simulators