Modeling subgrid-scale force and divergence of heat flux of compressible isotropic turbulence by artificial neural network

Toward neural-network-based large eddy simulation: application to turbulent channel flow

“…2019; Xie et al. 2019 a , b , c , 2020 a , b ) hidden layers, and Gamahara & Hattori (2017) showed that 100 neurons per hidden layer were sufficient for the accurate predictions of for a turbulent channel flow in a priori test. We also tested NN with three hidden layers, but more hidden layers than two did not further improve the performance both in a priori and a posteriori tests (see the Appendix).…”

“…In the case of compressible isotropic turbulence, Xie et al. (2019 a ) used FCNNs to predict SGS force and divergence of SGS heat flux, respectively, with the inputs of , , , , and at multiple grid points, where is the fluid density, and and are the mass-weighting-filtered velocity and temperature, respectively. Xie et al.…”

“…They (Xie et al. 2019 a , b , c , 2020 b ) showed that the FCNN-based LES provided more accurate kinetic energy spectrum and structure function of the velocity than those based on DSM and dynamic mixed model.…”

“…Zhou et al (2019) reported that using the filter size as well as the velocity gradient tensor as the input variables was beneficial to predict the SGS stresses for the flow having a filter size different from that of trained data. Xie, Wang & E (2020a) used an FCNN to predict the SGS force with the input of ∇ū at multiple grid points, and this FCNN performed better than DSM for the prediction of energy spectrum. In the case of three-dimensional decaying isotropic turbulence, Wang et al (2018) adopted the velocity and its first and second derivatives for the input of FCNN to predict the SGS stresses, and showed better performance in a posteriori test than that of DSM.…”

Artificial neural network subgrid models of 2D compressible magnetohydrodynamic turbulence

“…In our case, the outputs are all unique components of the desired SGS tensor which vary depending on the tensor of interest. Thus, we have O = 3 for τ kin and τ mag , O = 1 for τ ind , and O = 2 for τ enth This differs from most of the literature where a different ANN is used to find each individual component of the SGS tensor [19,[21][22][23]. By computing all components of the SGS tensor, we hope to incorporate physical symmetries into future models of τ AN N such as Galilean invariance, though we do not attempt to do so in this work.…”