What are the best practices for validating CFD and machine learning models for combustion applications?
Combustion is a complex physical and chemical process that involves high temperatures, turbulent flows, and multiple species. Computational fluid dynamics (CFD) is a powerful tool to simulate combustion phenomena and optimize the design and performance of engines, turbines, and reactors. However, CFD models often require high computational resources and rely on assumptions and simplifications that may introduce errors and uncertainties. Machine learning (ML) is an emerging technique that can complement CFD by providing data-driven insights, reducing computational costs, and enhancing predictive capabilities. But how can we ensure that CFD and ML models are reliable and accurate for combustion applications? In this article, we will discuss some of the best practices for validating CFD and ML models for combustion, based on the latest research and industry standards.