Studies on Vertical Axis Wind Turbines Technologies

本文将研究达里乌斯型垂直轴风力涡轮机（VAWT）的气动弹性和气动声学特性。Darrieus风力涡轮机在20世纪70年代很常见，但由于过度振动的问题，它们逐渐被淘汰。“智能”控制系统和叶片螺距控制的进步，以抑制振动和复合叶片结构的出现，使这种概念设计重新焕发活力。此外，为了尽量减少人口密集地区附近不可避免的视觉和噪音污染，需要在更深的水域建造海上风力发电场，这促使业界考虑漂浮式风力涡轮机。由于其固有的稳定性问题，水平轴风力涡轮机不适合漂浮装置。在HAWT中，由于重型发电机的作用，涡轮机的重心被向上推。在浮动装置中，重心需要靠近水面，以保证涡轮机的浮动稳定性。在这种情况下，VAWT是合适的，因为VAWT发电机可以位于涡轮机的底部。“troposkien”（来自希腊语的“旋转绳”）几何形状通常被用于设计Darrieus VAWT。“troposkien”是链条以一定速度旋转时的形状。在这种情况下，弯矩不存在，张力是结构中存在的唯一内力。因此，“troposkien”形状最大限度地减少了Darrieus几何的疲劳问题。考虑叶片结构曲率和叶片方位角区域气动失速的非线性效应，研究了非定常空气动力作用下“对流层”弯曲梁的结构动力特性。结构动力学将以状态向量混合公式形式表示，其中平面内外方向的力和位移将被视为因变量。VAWT的空气动力学模拟将结合自由尾流涡模型进行开发，其中每个叶片部分将由于叶片升力随时间的变化而产生涡。在拉格朗日流的描述中，由邻近涡旋的涡度和自由风速引起的速度场，将让脱落的涡旋（细丝或粒子）在空间上演化。塔（钝体）效应将被包括在内，以便更好地描述流。这种数值方法计算效率高，可以很好地表示旋转机翼中的自由尾迹，因为它是无混流的，并且对于传统（基于网格的）计算流体动力学方法中通常遇到的不现实的涡强度数值耗散具有很强的鲁棒性。本文将对非定常动力和气动载荷作用下结构的自激（颤振和极限环振荡）和气动弹性强迫响应进行数值研究。叶片上的非定常声压产生的气动声发射也将由叶片的局部升力来确定，以估计涡轮的噪声足迹，并与带有应变片的比例涡轮风洞试验数据进行比较。该项目的最终目标将是开发“智能”控制系统，以最大限度地减少在早期安装的Darrieurs转子中发现的高振动问题，并提高其在风力发电场安装中的性能（即能量提取），其中一台涡轮机的自由尾流受到其他涡轮机的影响。

The aeroelastic and aero-acoustic behaviour of Darrieus-type vertical axis wind turbines (VAWT) individually and in wind farm formations will be investigated. Darrieus wind turbines were common in the 1970's but they were phase out due to problems with excessive vibrations. Advances in "smart" control systems and blade pitch control to suppress vibrations and the advent of composite blade structures have revitalized this conceptual design nowadays. In addition, the necessity of building offshore wind farms in deeper water to minimize the unavoidable visual and noise pollution near populated areas have pushed the industry to consider floating wind turbines. Horizontal axis wind turbines are not appropriate to floating installations due to its inherent stability problems. In HAWT the centre of gravity of the turbine is pushed upwards due to the heavy electric generator. In floating installations, the centre of gravity needs to be close to the surface of the water to guarantee the floating stability of the turbine. VAWT are appropriate in this case because VAWT generators can be located at the base of the turbine. The "troposkien" (from the Greek "rotating rope") geometry has been usually invoked to design Darrieus VAWT. The "troposkien" is the shape that a chain takes when it is rotating at a certain rate. In this situation, the bending moment is absent and the tension is the only internal force present in the structure. Therefore, the "troposkien" shape minimizes the fatigue problems of the Darrieus geometry. The structural dynamic behavior of a curved beam shaped in the "troposkien" under the action of the unsteady aerodynamics will be investigated considering the non-linear effects due to both the structural curvature of the blade and the aerodynamic dynamic stall present in regions of the blade azimuth angle. The structural dynamics will be expressed in a state-vector mixed formulation form where both forces and displacements in the in- and out-plane directions will be considered as the dependent variables. The aerodynamic simulation of the VAWT will be developed incorporating a free-wake vortex model where each blade section will shed vortices due to the changes in the blade lift over time. The shed vortices (filaments or alternatively particles) will be let to evolve over space in a Lagrangian description of the flow by the velocity field induced by both the vorticity of the neighbouring vortices and the wind free speed. Tower (blunt-body) effects will be included for a better description of the flow. This numerical approach is computationally efficient to represent the free-wake in rotating wings as it is mash-less and so robust to unrealistic numerical dissipation of the vortex strength normally encountered in traditional (grid-based) computational fluid dynamics methods. Coupling the aforementioned models for the structure and the aerodynamics both the self-excited (flutter and limit cycle oscillations) and the aeroelastic forced response of the structure due to the unsteady dynamic and aerodynamic loading will be investigated numerically. Aero-acoustic emissions generated by the unsteady acoustic pressure on the blades will also be determined from the blade local lift to estimate the turbine noise footprint and compared with experimental data from wind tunnel tests with a scaled turbine instrumented with strain gauges. The final objective of the project will be to develop "smart" control systems to minimize the high vibration problems detected with earlier installations of the Darrieurs rotor and to enhance its performance (i.e. energy extraction) in wind farm installations where the free-wake of one turbine is affected by the presence of others.