ENHANCING AEROMECHANICAL ANALYSIS AND DESIGN CAPABILITIES OF WIND TURBINE ROTORS BY MEANS OF NONLINEAR FREQUENCY-DOMAIN COMPUTATIONAL FLUID DYNAMICS

在技术上和经济上可行的可再生能源中，风力发电是近年来开发增长最快的能源。这项研究的重点是现代水平轴风力涡轮机(HAWT‘s)，它通常具有两个或三个叶片的转子。风力机叶片的跨度从几米到100多米不等，其设计是一项复杂的多学科任务，需要考虑气动力和结构力的强烈非定常相互作用。气动不稳定的一些最危险的来源是a)由于风和旋翼平面的暂时非正交性而引起的偏航风，以及b)叶片动态失速。这些现象导致叶片受到时变的气动力，这可能会激发不希望看到的结构振动。这种情况反过来又会显著降低叶片及其支撑结构的疲劳寿命，从而导致过早的机械故障。这类事件可能会损害安装的技术和财务成功，这在很大程度上依赖于在10至30年的时间尺度上实现最低限度的服务的期望。这些事实突出了风力机叶片气动弹性设计过程的重要性。确定结构响应所需的非定常气动载荷必须在涡轮机的开发阶段被理解和准确地量化。由于所涉及的尺寸，在大多数情况下，无论是从经济角度还是从逻辑角度来看，进行气动弹性测试都是不可行的。因此，这些气动弹性问题只能通过精确的模拟工具来解决。这个项目的总体动机是双重的：它的目的既是为了丰富与风力机气动弹性相关的非定常流动知识，又是为了推动最先进的计算技术来完成这一任务。这些目标是通过使用一种新的计算流体动力学(CFD)方法来研究风力机的非定常空气动力学来实现的。采用非线性频域(NLFD)技术求解三维非定常粘性流动方程，确定了与气动弹性分析相关的非定常周期流动。NLFD-CFD方法已成功地应用于固定翼和叶轮机械气动弹性问题。这项研究将利用这种高保真的方法来加强对HAWT叶片严重非定常气动强迫的理解，并大大降低相对于传统的时域CFD分析的计算成本。这种方法特别适合于研究失速引起的振动和偏航风致的非定常气动叶片载荷。另一方面，这项技术将极大地帮助设计者开发新的叶片，而不需要依赖现有的翼型数据数据库，而目前的大多数分析和设计系统都依赖于这些数据库。该项目的主要成果之一将是大大减少风力机非定常空气动力学和气动弹性三维非定常粘性流动模型的物理精度和计算负担能力之间的矛盾。这项研究的成果将使英国和欧洲的工业受益，因为它们将为设计更高效和更可靠的叶片提供有效的工具。NLFD-CFD技术还将提供对影响风力涡轮机疲劳寿命的非定常气动现象的更深入了解。在接下来的几年里，风力涡轮机的认证过程将对该行业实施更严格的要求。开发的技术将支持满足增强认证标准所需的分析。非定常空气动力学研究界作为一个整体也将从这项研究中受益，因为它的发现将加强和巩固NLFD技术在旋翼飞行器、涡轮机械和飞机气动弹性方面的部署。

Among technically and economically viable renewable energy sources, wind power is that which exploitation has been growing fastest in the recent years. This research focuses on modern Horizontal Axis Wind Turbines (HAWT's), which typically feature two- or three-blade rotors. The span of HAWT blades can vary from a few meters to more than 100 meters, and their design is a complex multidisciplinary task which requires consideration of strong unsteady interactions of aerodynamic and structural forces. Some of the most dangerous sources of aerodynamic unsteadiness are a) yawed wind, due to temporary non-orthogonality of wind and rotor plane, and b) blade dynamic stall. These phenomena result in the blades experiencing time-varying aerodynamic forces, which can excite undesired structural vibrations. This occurrence, in turn, can dramatically reduce the fatigue life of the blades and their supporting structure, yielding premature mechanical failures. Events of this kind can compromise the technical and financial success of the installation, which heavily relies on fulfilling the expectations of minimal servicing on time-scales of the order of 10 to 30 years. These facts highlight the importance of the aeroelastic design process of HAWT blades. The unsteady aerodynamic loads required to determine the structural response must be understood and accurately quantified in the development phase of the turbine. Due to the sizes at stake, in most cases it is infeasible to perform aeroelastic testing, not only from an economic but also logistic viewpoint. Hence these aeroelastic issues can only be tackled by using accurate simulation tools.The general motivation of this project is two-fold: it aims both at enriching the knowledge of unsteady flows relevant to wind turbine aeroelasticity, and advancing the state-of-the-art of the computational technology to accomplish this task. These objectives are pursued by using a novel Computational Fluid Dynamics (CFD) approach to wind turbine unsteady aerodynamics. The unsteady periodic flow relevant to aeroelastic analyses is determined by solving the three-dimensional unsteady viscous flow equations with the nonlinear frequency-domain (NLFD) technology. The NLFD-CFD approach has been successfully applied to fixed-wing and turbomachinery aeroelasticity. This research will exploit this high-fidelity methodology to enhance the understanding of the severe unsteady aerodynamic forcing of HAWT blades, and substantially reduce computational costs with respect to conventional time-domain CFD analyses. This method is particularly well suited to investigate the unsteady aerodynamic blade loads associated with stall-induced vibrations and yawed wind. On the other hand, this technology will greatly help designers to develop new blades without relying on the database of existing airfoil data on which the majority of present analysis and design systems depend. One of the main results of this project will be to greatly reduce the dichotomy between the conflicting requirements of physical accuracy and computational affordability of the three-dimensional unsteady viscous flow models for wind turbine unsteady aerodynamics and aeroelasticity. The achievements of this research will benefit the British and European industry in that they will offer an effective tool to design more efficient and reliable blades. The NLFD-CFD technology will also provide deeper insight into unsteady aerodynamic phenomena which affect the fatigue life of wind turbines. In the next few years, the certification process of wind turbines will enforce stricter requirements on the industry. The developed technology will support the analyses required to meet enhanced certification standards. The Unsteady Aerodynamics Research Community as a whole will also benefit from this research, because its findings will enhance and consolidate the deployment of the NLFD technology in rotorcraft, turbomachinery, and aircraft aeroelasticity.