Neural Prediction Model for Transition Onset of a Boundary-Layer in Presence of 2D Surface Defects

Abstract

Predicting the laminar to turbulent transition is an important aspect of computational fluid dynamics because of its impact on skin friction. Traditional methods of transition prediction do not make it possible to consider configurations where the boundary layer develops in presence of surface defects (bumps, steps, gaps, etc.). A neural network approach is used in this paper, based on an extensive database of boundary layer stability computations in presence of gap-like surface defects. These computations consists on linearized Navier-Stokes calculations and provide informations on the effect of surface irregularity geometry and aerodynamic conditions on the transition to turbulence. Physical and geometrical parameters characterizing the defect and the flow are provided to a neural network whose outputs inform about the effect of a given gap on the transition through the Δ𝑁 method.

Neural prediction model for transition onset of a boundary layer in presence of two-dimensional surface defects

“…Supplementary material. The three NNs developed in this study are available at the following address: https://doi.org/10.5281/ zenodo.7101195 (Rouviere, Pascal, Méry, Simon, & Gratton, 2022). The repository also includes a Python script named Main.py and a Jupyter script named Main.ipynb that provide detailed explanations and examples on how to use these NNs.…”

Laminar-turbulent transition experiment on the effect of surface imperfections on an NLF airfoil in transonic flow

“…the step component have to be considered. Figure 13a More details on the numerical method used to perform the stability calculations can be found in Rouviere et al (2022).…”