Neural Network for Subgrid Turbulence Modeling for Large Eddy

Abstract: When simulating multiscale systems, where some fields cannot be fully prescribed despite their effects on the simulation's accuracy, closure models are needed. This behavior is observed in turbulent fluid dynamics, where Large Eddy Simulations (LES) depict global behavior while turbulence modeling introduces dissipation correspondent to smaller sub-grid scales. Recently, scientific machine learning techniques have emerged to address this problem by integrating traditional (physics-based) equations with data-driven (machine-learned) models, typically coupling numerical solvers with neural networks. This work presents a comprehensive workflow, encompassing high-fidelity data generation, a priori learning, and a posteriori learning, where data-driven models enhance differential equations. The study underscores the critical role of post-processing and effective filtering of fine-resolution fields and the implications numerical methods selection, such as the Lattice Boltzmann Method (LBM) or Finite Volume Method.