厚胶层FRP粘接结构的超声反射谱测量及粘接质量表征方法研究

目前，具有厚胶层特点的FRP（Fiber Reinforced Polymer，纤维增强复合材料）粘接结构在大型、轻质、高强度构件的制造方面取得了广泛应用，较典型的例子为风力发电领域中的风机叶片。因此，针对厚胶层FRP粘接结构粘接质量的无损检测对风机叶片在制造及服役过程中的完整性评价意义重大。本项目基于超声反射谱测量方法，拟开展厚胶层FRP粘接结构粘接质量量化表征的方法研究。主要内容包括：（1）建立厚胶层FRP粘接结构的等效粘接模型，提出一种求解超声反射特性的新思路，研究胶层及粘接界面性能参数对超声反射特性的影响机理；（2）结合超声反射特性的理论分析，研究厚胶层FRP粘接结构中不同界面处超声反射系数谱的测量方法；（3）围绕超声反射谱中的特征参数、胶层及粘接界面性能参数与粘接质量之间的内在联系，建立基于人工神经网络识别算法的粘接质量预测模型，实现厚胶层FRP粘接结构粘接质量的量化表征。

Currently, FRP (Fiber Reinforced Polymer) bonding structures with thick adhesive layer are widely used in the manufacture of large, lightweight and high-strength components. The typical applications of this kind of bonding structure are wind turbine blades in the field of wind power generation. Therefore, nondestructive testing for the bonding quality of FRP bonding structures with thick adhesive layer is of great significance for integrity evaluation both in the manufacturing and service process of wind turbine blades. Based on measurement of ultrasonic reflection spectrum, this project intends to carry out the quantitative characterization for bonding quality of FRP bonding structures with thick adhesive layer. The main contents include three aspects. Firstly, an equivalent bonding model for analyzing the ultrasonic reflection characteristics of FRP bonding structures with thick adhesive layer will be established, and a new way to solve the ultrasonic reflection spectrums will be introduced. Based on these, the influencing mechanism of the properties of adhesive layer and bonding interface on the ultrasonic reflection characteristics will be systematically studied as well. Secondly, combined with the theoretical analysis of ultrasonic reflection characteristics, the measuring method of ultrasonic reflection coefficient spectrum at different interfaces of FRP bonding structures with thick adhesive layer will also be investigated. Finally, according to the internal relationship among the characteristic parameters (resulting from ultrasonic reflection spectrum, the adhesive layer and the bonding interface) and bonding quality of the bonding structure, a prediction model based on artificial neural network identification algorithm will be set up to realize the quantitative characterization of the bonding quality of FRP bonding structures with thick adhesive layer.