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

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 offline-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., 2022 [8]) 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.深度学习领域的最新进展为科学计算社区带来了诸多启发，其被视为加速计算密集型反应流模拟的一种极具前景的方法。本研究介绍了DeepFlame——一款开源C++平台，可借助机器学习算法与离线训练模型求解反应流问题。本平台融合了计算流体动力学库OpenFOAM、机器学习框架Torch以及化学动力学程序Cantera的各自优势。作为DeepFlame核心的跨库函数与数据交互复杂度被降至最低，以实现代码维护、扩展与升级的简洁清晰工作流。作为演示，我们将此前用于预测化学动力学的深度学习研究（Zhang等人，2022年[8]）应用于此，以凸显机器学习在加速反应流模拟中的潜力。本研究通过大量典型算例开展了全面的代码验证，以评估其准确性与计算效率。结果表明，DeepFlame中实现的对流-扩散-反应算法在稳态与瞬态过程中均具备出色的鲁棒性与准确性。此外，本研究还探索了多种旨在进一步提升计算效率的方法，例如动态负载均衡与自适应网格细化，并对这些方法的性能进行了评估与报告。借助本研究实现的深度学习方法，在中端图形处理器（Graphics Processing Unit，GPU）上运行的简单氢点火算例中，实现了两个数量级的加速。针对碳氢燃料与其他复杂燃料的模拟，计算效率有望获得进一步提升。在人工智能专用芯片——深算单元（Deep Computing Unit，DCU）上也实现了相近量级的加速，凸显了DeepFlame在适配下一代计算架构与硬件方面的潜力。