This study aims to investigate the performance of a data-driven methodology for quantifying damage based on the use of a metamodel obtained from the Polynomial Chaos-Kriging method. The investigation seeks to quantify the severity of the damage, described by a specific type of debonding in a wind turbine blade as a function of a damage index. The damage indexes used are computed using a data-driven vibration-based structural health monitoring methodology. The blade’s debonding damage is introduced artificially, and the blade is excited with an electromechanical actuator that introduces a mechanical impulse causing the impact on the blade. The acceleration responses’ vibrations are measured by accelerometers distributed along the trailing and the wind turbine blade. A metamodel is formerly obtained through the Polynomial Chaos-Kriging method based on the damage indexes, trained with the blade’s healthy condition and four damage conditions, and validated with the other two damage conditions. The Polynomial Chaos-Kriging manifests promising results for capturing the proper trend for the severity of the damage as a function of the damage index. This research complements the damage detection analyses previously performed on the same blade.

这项研究的目的是基于从多项式混沌克里格方法获得的元模型的使用，研究数据驱动的方法来量化损害的性能。该研究试图量化损坏的严重程度，该损坏的严重程度由风力涡轮机叶片中的特定类型的脱胶描述为损坏指数的函数。使用的损坏指数是使用基于数据驱动的基于振动的结构健康监测方法来计算的。刀片的脱粘损坏是人为引入的，而刀片则由机电致动器激励，该机电致动器引入了机械冲击，从而对刀片产生冲击。加速度响应的振动是通过沿尾部和风力涡轮机叶片分布的加速度计进行测量的。以前，通过基于损坏指数的多项式混沌克里格方法获得了元模型，并使用叶片的健康状况和四个损坏状况对其进行了训练，并使用其他两个损坏状况对其进行了验证。多项式混沌克里格法显示出令人鼓舞的结果，可以捕捉到随损伤指数而定的损伤严重程度的正确趋势。这项研究是对先前在同一刀片上执行的损伤检测分析的补充。