Speaker
Description
Turbulence in fluids represents one of the most complex and challenging phenomena in fluid mechanics. Accurately modeling turbulence is crucial for various engineering and scientific applications, such as aerodynamics, meteorology, and environmental engineering. Traditional numerical methods, including Direct Numerical Simulations (DNS) and Large Eddy Simulations (LES), demand substantial computational resources and have limitations in scale and applicability.
This poster aims to review the current machine learning (ML) and artificial intelligence (AI) techniques used for modeling turbulence. Beginning by outlining the fundamental nature of turbulence and the inherent difficulties in its modeling, including its chaotic behavior, high Reynolds number flows, and the wide range of spatial and temporal scales involved. Understanding these challenges highlights the importance of developing more efficient and accurate modeling methods.
Furthermore, there is an exploration of various ML and AI approaches that have been proposed to tackle these challenges. Techniques such as deep learning (DL), neural networks (NN), and unsupervised learning algorithms are used to predict turbulent structures and their dynamics. These methods offer significant advantages in processing large datasets from experiments and simulations, leading to the development of more precise and efficient models.
This review aims to present the benefits and limitations of these approaches compared to traditional methods, emphasizing the potential of ML and AI to revolutionize turbulence modeling. Integrating these techniques with conventional numerical methods can result in hybrid models that better predict and control turbulent flows. This advancement holds promise for enhanced performance in various practical applications, paving the way for future research and development.
