2.5D large eddy simulation of vertical axis wind turbine in consideration of high angle of attack flow

Abstract

A renewed interest in vertical axis wind turbines (VAWTs) has been seen recently. Computational fluid dynamics (CFD) is regarded as a promising technique for aerodynamic studies of VAWTs. In particular, 2D unsteady Reynolds-averaged Navier–Stokes (URANS) is commonly adopted, although past studies on VAWTs revealed the limited accuracy of 2D URANS. This paper investigated the feasibility and accuracy of three different CFD approaches, namely 2D URANS, 2.5D URANS and 2.5D large eddy simulations (LES), in the aerodynamic characterization of straight-bladed VAWT (SBVAWT), with a focus on the capability of the 2.5D LES approach in CFD simulation of high angle of attack (AOA) flow. The 2.5D model differs from a full 3D model in that only a certain length of blades is modeled with periodic boundaries at the two extremities of the domain. The applications of these three approaches were systematically examined in the aerodynamic simulations of a single static airfoil and a 3-blade SBVAWT at different rotating speeds. Their capabilities to predict the aerodynamic forces were evaluated through a comparison with the wind tunnel results obtained by other researchers, with particular attention to high AOA flow beyond stall. Among the three methods, 2.5D LES yielded the best agreement with the experimental results in both cases. Detailed examinations of simulated flow field revealed that 2.5D LES produces more realistic 3D vortex diffusion after flow separation, resulting in more accurate predictions of aerodynamic coefficients in static or dynamic stall situations. It is noteworthy that 2.5D LES cannot capture the effect of tip vortex and vertical flow divergence in VAWTs, which used to be regarded by some researchers as the major cause of overprediction of VAWT power in 2D URANS. In this study, the considerably improved results achieved by 2.5D LES imply that the poor accuracy of URANS method is mainly due to its inherent limitation in vortex modeling. In general, 2.5D LES showed good agreement with experimental results at a relatively low tip speed ratio (TSR), but only fair agreement at a high TSR. Compared with the other two approaches, 2.5D LES is regarded as a more promising and effective CFD tool for investigating the aerodynamic characteristics of VAWTs, particularly their self-starting features corresponding to very low rotation speeds.