Convolutional neural network models and interpretability for the anisotropic reynolds stress tensor in turbulent one-dimensional flows

The Reynolds-averaged Navier-Stokes (RANS) equations are widely used in turbulence applications. They require accurately modelling the anisotropic Reynolds stress tensor, for which traditional closure models only yield reliable results some flow configurations. In last few years, there has been a surge of work aiming at using data-driven approaches to tackle this problem. majority previous focused on development fully connected networks tensor. paper, we expand upon recent turbulent channel and develop new convolutional neural network (CNN) that able predict normalised We apply CNN model number one-dimensional flows. Additionally, present interpretability techniques help drive design provide guidance behaviour relation underlying physics.