复杂环境荷载下海洋单桩风电机的随机响应与破坏机理分析

By 2030 the majority of wind farm developments in China will occur in offshore Eastern and Southeastern regions and the monopile foundation will continue to be the dominant structural style of wind turbines to be employed there. Considering that the wind farms in these areas are earthquake prone and subject to harsh marine environments, it is of utmost importance to investigate the stochastic behavior of offshore monopile wind turbines and their failure mechanism under the excitation of wind, wave, current and earthquake actions. This project comprises the following four state-of-the-art research topics: (1) Dynamic analyses and laboratory tests of wind turbine structural response under simultaneous seismic and wave excitations (2) Seismic resistant capacity of wind turbines using Pushover analysis (3) Extreme response prediction of wind turbines in survival sea conditions (4) Long-term nonlinear fatigue damage assessment. The earthquake shaking table to be used is in a wave tank. It is so designed that both seismic excitation and ocean waves can be generated and can act simultaneously on an offshore structure. This new idea of including waves physically in the shaking table tests was actually not realized by researches before. Also the governing equation of structural motions is solved. The comparative study between experiments and calculations thus will help to reveal the failure mechanism of monopile wind turbines under both hydrodynamic and seismic loads. Another approach of the seismic study is Pushover analysis which was rarely applied in the literature for offshore wind turbines. An improved decoupling adaptive spectral method is proposed in this project to further investigate the earthquake resistant capacity of wind turbine structures. For predicting the extreme structural response of offshore monopile wind turbines in a survival condition, the nonlinear stochastic waves in shallow water and the transformation with seabed slope, as well as nonlinear wave forces, are taken into account. Time-domain wave simulations adopt two methods: Boussinesq formulation and, higher-order nonlinear stochastic wave theories combining wave spectrum transformation. Due to lengthy computation and stochastic uncertainty in simulations, a frequency-domain analysis method is therefore developed to efficiently capture the extreme response. This method is based on formulating a multi-input/multi-output Volterra series model and on cumulant spectral analysis. The long-term fatigue damage assessment shares some similarities with this frequency-domain approach, whereas the differences lie in the long-term joint distribution of wind speed and wave height and the stochastic property of wind load that makes the stress response to be wide-banded. For this we examine the independence of wind and wave processes and apply this property to the bi-spectral analysis of wind turbine structural response for an efficient nonlinear fatigue assessment.

我国海上风电场建设于2030年前仍将以浅海中的单桩基础风电机为主，多处位于华东和东南沿海等海洋环境多变、地震易发的区域。本项目着眼于研究近海单桩风机在复杂的风浪流海况与地震下的结构随机响应与破坏机理。研究主要包括风机在有浪下的地震响应分析与振动台试验、风机在极端海况下结构的非高斯响应极值、风机结构的长期非线性随机疲劳、Pushover抗震性能研究这四个关键科学问题，采用理论分析、数值模拟和模型试验相结合的研究思路。地震振动台位于波浪水池中，可模拟地震激励和波浪荷载同时加载，这是一大创新。再通过理论分析与计算对比,揭示风机在浪与地震共同作用下的破坏机理。我们还提出完全解耦的适应谱Pushover法应用到抗震性能研究中。极端海况下的结构响应极值估计，考虑到了浅水中波浪的非线性、随机性及波浪力的非线性，运用时域模拟与基于文特拉级数和累积谱的频域分析。非线性随机疲劳分析则包括了风荷载与系统非线性。

我国海上风电场建设仍将以浅海中的单桩基础风电机为主，多处位于华东和东南沿海等海洋环境多变、地震易发的区域。本项目着眼于研究近海单桩风机在复杂的风浪流海况与地震下的结构随机响应与破坏机理。研究主要包括风机在有浪下的地震响应分析与振动台试验、风机在极端海况下结构的非高斯响应极值、风机结构的长期非线性随机疲劳、Pushover抗震性能研究这四个关键科学问题，采用理论分析、数值模拟和模型试验相结合的研究思路。其中，海上单桩风机在地震和波浪联合作用下的响应研究在世界上领先，我们开展了1/30模型试验，试验工况多达30多种，通过试验，我们获得诸多重要发现，比如共振、地震波浪耦合以及一个有趣而重要的现象，即whirling vibration涡旋振动，这是一个非线性振动的国际性难题。对于海上单桩风机的结构非高斯响应极值，本研究创新性地提出了将中心矩C-moments和线性矩L-moments联合使用的方法，因为后者对样本长度不敏感。若使用峰度和5阶线性矩，极值估计的精度与鲁棒性极好。此外，我们还研究了极端海况下波浪在多腿单桩风机上的爬升与砰击的试验和计算流体模拟研究，一个重要发现是绕圆桩的最小爬升发生在近柱后的145度，而非之前文献中所载的135度，这为实际海上的维修船人员登陆风电机以及爬梯的前期设计提供了很有意义的参考。多桩风机上爬升系数和砰击系数和单桩存在着明显的不同，需要依照具体情形予以考虑。对于结构非线性随机疲劳分析，创新性地引入权重系数对低频处的功率谱密度进行修正，以考虑高频对低频的影响，该法和传统的JM法、TB法、Low法相比，鲁棒性更强、精确度更高、适用性更广。本项目在论文发表和专利申请上取得了不俗的成果，共有11篇论文已发表，1篇正刊印中，这包括4篇SCI，2篇EI论文。所有SCI论文都是JCR-2区以上。此外，还有4项专利得以授权。本项目还培养了5名硕士研究生、2名博士研究生、2名博士后。