The wind energy sources promote the generation of electricity, being environmentally safe, clean and its use has been growing in Brazil, especially in the northeastern states with the implementation of new wind farms. Wind turbines, however, are subject to weather, risk of animal shocks, materials carried by the wind, and the vibrations of the system itself. The blade is the most important component of a wind turbine and is the one with the greatest risk of failure. They are usually made of a composite fiber-reinforced polymer matrix and balsa wood as structural reinforcement. As the material composing this blade has heterogeneous and anisotropic characteristics, conventional inspection techniques are not considered effective. The non-destructive Tap Test technique can be a safe option in composite materials because it does not suffer from these limitations. The objective of this work is to inspect composite plates of polymeric resin used in wind turbines with discontinuities using the non-destructive Tap Test technique, where regions with and without defect were analyzed. The collected signals from an accelerometer and a microphone were processed, to allow the extraction of features and the recognition of discontinuities with the aid of a neural network.

风能资源促进了环境安全，清洁的发电，在巴西，尤其是在东北部各州，随着新风电场的兴起，其使用量一直在增长。但是，风力涡轮机会受到天气，动物撞击风险，风携带的材料以及系统本身的振动的影响。叶片是风力涡轮机最重要的组成部分，也是故障风险最大的部分。它们通常由复合纤维增强的聚合物基体和轻木作为结构增强材料制成。由于构成该叶片的材料具有异质性和各向异性特征，因此常规的检查技术被认为无效。在复合材料中，无损抽头测试技术可能是安全的选择，因为它不受这些限制的影响。这项工作的目的是使用非破坏性抽头测试技术检查用于风力涡轮机的具有不连续性的聚合树脂复合板，其中分析了有缺陷和无缺陷的区域。处理从加速度计和麦克风收集到的信号，以便借助神经网络提取特征并识别不连续性。