Computational Combustion
and Machine Learning (CCML) Lab 

At the fundamental level, the CCML Lab focuses on developing generalizable methods to model essential phenomena in turbulent combustion, such as chemical kinetics, turbulence–chemistry interactions, hardware acceleration techniques (CPU+GPU integration), and complex geometries using data-driven methods. Our objectives include designing robust machine learning architectures for efficient training, validating these models against experimental data from real engines, and integrating them with open-source/commercial CFD codes. By maintaining strict physical fidelity while enhancing computational efficiency, we aim to push the boundaries of high-fidelity simulation without compromising on practical runtimes.

On the applied side, the CCML Lab plans to target a wide spectrum of combustion systems, including those in aviation propulsion, power generation, and ground transportation. We intend to investigate flame structures, forecast pollutant formation, and analyze system stability with rapid, high-fidelity modeling solutions - key for industrial design and environmental compliance. By merging high-performance computing with advanced data-driven methodologies, our goal is to develop reduced-order models that capture the essential physics while significantly cutting computational costs, thereby opening new avenues for clean and efficient combustion technologies.