| The wind turbine works in the atmosphere boundary layer. There the flow field is usually three-dimensional (3D) and unsteady, which makes it very difficult to analyze the3D flow field and the corresponding aerodynamic performance of the Horizontal Axis Wind Turbine (HAWT). Besides, as the largest rotating turbomachinery on the earth, the aerodynamic performance of HAWT (102m magnitude) has very close relationship with the boudary layer on the wind turbine blade and the structure of the shedding vortexes, whose magnitude is only about10"3m. From the macro large scale to the micro small scale, the multiple scale structures of the3D flow field present new challenges to the3D flow field study and the3D aerodynamic research of the HAWT. According to the computational complexity and the accuracy differencies, the existing aerodynamic computational models of HAWT can be mainly divided into three parts. They are the Blade Element and Momontum (BEM) model, Vortex Method Model and the Computational Fluid Dynamics (CFD) model. Because these models have their own independent characters and the computational features for the3D flow field analysis, they have different scopes of application.In this paper, in order to improve the3D computational ability of the current BEM Model, Vortex Method Model and the CFD Model, the following works are carried out:Firstly, for the BEM model part, the3D stall delay modification model is mainly studied. The analytical solution of the Inviscid Stall Delay Model, which is proposed by Cotern, is derived. Based on the solution, the3D Inviscid Stall Delay Modification (ISDM) Model is created. In this model, we treat the stall delay effects differently by the delay of the separation point on the airfoil, and aim to capture the further negative pressure reduction in the separation area. By simplifying the Navier-Stokes equations and introducing the Kirchhoff-Helmholz trailing edge separation prediction model, the3D stall delay effect can be evaluated under the influence of the Centrifugal and Coriolis forces. At last, the ISDM model is validated by the full scale3D wind tunnel experimental results of the NREL Phase VI and MEXICO wind turbines, which verified the accuracy of the ISDM model.Secondly, for the Vortex Method Model part, the Lifting-Line Model and the3D Panel-Method Model, which have different forms of the vorticity representation and the computational complexity, are detailed studied. In the Lifting-Line Model, by introducing the Viscous Vortex Core model and the vortex effective radius concept, the numerical singularity can be effectively resolved and at the same time the accuracy of the3D unsteady free wake model can be improved. Based on the Lifting-Line Model, the aerodynamic performance and the wake characters of the backward-swept wind turbine blade are anaylized. In the Panel-Method Model, the2D boundary layer computational model is involved and based on the Direct Coupling Strategy a simple Viscous and Inviscid Interaction (VII) model is created. Compared with the results of the Panel Method Model, the VII model improves the accuracy of the computation under the attached flow condition. However when the Angle of Attack (AOA) becomes large enough and the flow separates from the surface, the VII model is difficult to get a convergence solution. Overall, because of the3D induced velocity of the3D wake vortex, the Vortex Model greatly improves the computational ability of the3D flow field.Thirdly, for the CFD model part, the study mainly focuses on the Actuator Model. Based the existing Actuator Disc Model (ADM), Actuator Line Model (ALM) and Actuator Surface Model (ASM), a new model named Improved Actuator Surface Model (IASM) is proposed, which is a kind of combination of the Actuator Model and the VII model. By making full use of the Boundary Element features of the VTI model, the IASM improves the interaction mechanism between the3D geometry of the wind turbine blade and the3D flow field to the maximum extent. By comparing the computatinoal results of the IASM and ALM, it is validated that the IASM can increase the accuracy of the flow field results in the near-wake region. The IASM provides an efficient way for the3D flow field computation and the aerodynamic analysis of the large scale wind turbine especially with complex geometry.Fourthly, for the experiment part, based on the0.5m X0.5m closed wind tunnul of the Institute of Engineering Thermophysics (LET), Chinese Academy of Sciences, a special swept airfoil section experiment is designed, which is aim to study the airfoil aerodynamic performance infuenced by the spanwise flow. In order to improve the reliability of the experimental data, a specialized data correction model proposed by Kang, which contains the blockage effect caused by the flow separation, is utilized in this dissertation. The swept airfoil experiment shows that the lift coefficient of the swept airfoil section is close to the results of the finite straight airfoil at small AOA condition, but at the stall stage, the lift coefficient of the swept airfoil section presents big-scale fluctuations and the average value of the lift coefficient is significantly higher than that of the straight airfoil section. Finally, based on the experimental results, the ISDM is detailed evaluaed. The results shows that the ISDM can accurately estimate the lift coefficient undert the attached flow and primary trailing edge separation conditions, but there are still errors exist in the drag coefficient estimated by ISDM. |