Abstract
This paper numerically investigates four methods, namely mesh refinement, General Richardson Extrapolation (GRE), Grid Convergence Index (GCI), and the fitting method, in order to obtain a mesh independent solution for a straight blade vertical axis wind turbine (SB-VAWT) power curve using computational fluid dynamics (CFD). The solution is produced by employing the 2D Unsteady Navier–Stokes equations (URANS) with two turbulence models (Shear Stress Transport (SST) Transitional and ReNormalized Groups (RNG) κ−ɛ models). The commonly applied mesh refinement is found to be computationally expensive and not often practical even for a full 2D model of the turbine. The mesh independent power coefficient produced using the General Richardson Extrapolation method is found to be encouraging. However, the Grid Convergence Index may not be applicable in mesh independency tests due to the oscillatory behaviour of the convergence for the turbine power coefficient. As an alternative, the fitting method shows a good potential for the predicting of the mesh independent power coefficient without the necessity to consider a massive number of meshes.
•Critically review the current status of the published CFD simulations on VAWT.•Validating the SST Transitional and RNG κ−ɛ model with the available experimental data.•Mesh independency test using mesh Resolution, General Richardson Extrapolation, Grid Convergence Index, Fitting method.•Employing fitting method in convergence analysis at different tip speed ratio.•Convergence oscillatory behaviour is observed in SB-VAWTs.