To main content

Numerical assessment of RANS turbulence models for the development of data driven Reduced Order Models

Abstract

The ability to accurately predict vortex shedding around wind turbine blades is paramount, particularly at high Reynolds number. Turbulence models employed in the numerical studies strongly influence flow separation and the aerodynamic loading, thus affecting the overall accuracy of numerical simulations. In this manuscript, three turbulence models (Spalart–Allmaras, and Shear Stress Transport model) are investigated in two and three-dimensional configurations using standard Reynolds Average Navier–Stokes equations. The focus is on the NACA0015 airfoil, and the simulations are conducted at a Reynolds number of to match the experimental data in the literature. The effect of flow separation and vortex shedding pattern is investigated at different angles of attack (0 17) along with the prediction ability of the turbulence models. Spectral analysis is performed over the time history of aerodynamic coefficients to identify the dominant frequencies along with their even and odd harmonics. A reduced-order model based on the van der Pol equation is proposed for the aerodynamic lift calculation. The method of multiple scales (a perturbation approach) is adapted to compute the coefficients of the proposed model consisting of quadratic and cubic nonlinearities at the various angle of attacks (). The model is also tested in a predictive setting, and the results are compared against the full order model solution.

Category

Academic article

Language

English

Author(s)

Affiliation

  • SINTEF Digital / Mathematics and Cybernetics
  • Norwegian University of Science and Technology

Year

2020

Published in

Ocean Engineering

ISSN

0029-8018

Volume

196

View this publication at Norwegian Research Information Repository