Overview
This course focuses on the potential of machine learning to enhance computational fluid dynamics. The learning outcomes include understanding how to accelerate direct numerical simulations, improve turbulence closure modeling, and develop enhanced reduced-order models. The course teaches skills such as learning data-driven discretizations for partial differential equations, enhancing shock capturing schemes via machine learning, and utilizing various models like RANS and LES. The teaching method involves a combination of lectures, practical examples, and incorporating physics into turbulent flow solvers. The intended audience for this course includes students and professionals in the fields of computational fluid dynamics, scientific computing, and machine learning.
Syllabus
Intro
ML FOR COMPUTATIONAL FLUID DYNAMICS
Learning data-driven discretizations for partial differential equations
ENHANCEMENT OF SHOCK CAPTURING SCHEMES VIA MACHINE LEARNING
FINITENET: CONVOLUTIONAL LSTM FOR PDES
INCOMPRESSIBILITY & POISSON'S EQUATION
REYNOLDS AVERAGED NAVIER STOKES (RANS)
RANS CLOSURE MODELS
LARGE EDDY SIMULATION (LES)
COORDINATES AND DYNAMICS
SVD/PCA/POD
DEEP AUTOENCODER
CLUSTER REDUCED ORDER MODELING (CROM)
SPARSE TURBULENCE MODELS
Taught by
Steve Brunton