A novel ANN-Based boundary strategy for modelingmicro/nanopatterns on airfoil with improved aerodynamic performances

In this paper, in order to overcome computational challenge in numerical simulations of airfoil covered with micro/nanopatterns for aerodynamic drag-reduction design, an ANN-Based boundary strategy is proposed to balance the accuracy and efficiency. Lattice Boltzmann Method (LBM) is utilized to extract the micro/nanoflow characteristics in the near-wall region considering the rarefied effect and the available microscopic data is used to train the surrogate boundary model by Generalized Regression Neural Network (GRNN). The modified boundary conditions obtained by ANN-Based model replacing the real complex and fine micro/nano structure are applied on the smooth configuration to perform macroscopic simulations. Rectangular riblets as micro/nano pattern are discussed for our study. Various arrangement strategies of rectangular riblets over airfoil are adopted by adjusting the width, height, and coverage area to investigate their effects on aerodynamic performances. The results indicate that for the riblet airfoil, the skin friction reduces and the transition position moves backward compared with those of the smooth airfoil. Furthermore, the lift-to-drag ratio also significantly increases and the rate of improvement is up to 13.8% at the Angle of Attack (AOA) of 1 degrees. This paper shows a perspective in aerodynamic design with micro/nano pattern for drag reduction by providing an innovative multi-scale simplified simulation strategy. (C) 2022 Elsevier Masson SAS. All rights reserved.

Automated summary

This paper proposes an ANN-based boundary strategy to tackle the computational challenge in numerical simulations of airfoils covered with micro/nanopatterns for aerodynamic drag reduction design. The lattice Boltzmann method is used to extract the micro/nanoflow characteristics, and a surrogate boundary model is trained using microscopic data and the generalized regression neural network. The results show that the rectangular riblet airfoil with micro/nano pattern exhibits reduced skin friction, backward transition position, and significantly increased lift-to-drag ratio.