Abstract: The twist angles of helicopter rotor blades had a direct impact on the performance and service life of the blades. However, the traditional methods of detecting twist angles relied heavily on specialized fixtures and measuring tools, which were inflexible, inefficient and inaccurate. To address this issue, a digital flexible detection method for rotor blade twist angles was proposed, and a detection system for rotor blade twist angles was designed and constructed. This system enabled automatic detection of rotor blade twist angles and might greatly improve detection efficiency and accuracy to meet the detection requirements of different types of rotor blades. A twist angle calculation method was proposed based on key section local contour registration. The point clouds on the surfaces of rotor blades were collected by laser scanning sensor, a point cloud registration method was employed by combining coarse registration based on principal component analysis(PCA) algorithm and fine registration based on improved iterative closest point using K-dimension tree(KD-ICP) algorithm to accurately and efficiently calculate the twist angle values, which provided a general and effective calculation method for rotor blade twist angle calculation. Finally, experiments were conducted on a certain type of rotor blade as the detection object. The results show that the maximum error of twist angle detection is as 54″, and the repeatability accuracy is as 18″, which proves the effectiveness and accuracy of the systems.

Key words: twist angle, rotor blade, digital detection, laser scanning, point cloud registration

摘要： 旋翼桨叶扭角直接影响着桨叶的性能和使用寿命，而传统扭角检测多依赖于专用的工装和量具，检测方法柔性度低、检测效率低、重复精度低。针对此问题，提出了旋翼桨叶扭角的数字化柔性检测方法，设计并构建了旋翼桨叶扭角检测系统，实现了旋翼桨叶扭角的自动化检测，可在较大程度上提高检测效率和精度，满足不同类型桨叶的检测需求。提出了一种基于关键截面局部轮廓配准的扭角计算方法，利用激光扫描传感器采集旋翼桨叶表面点云，采用基于主成分分析(PCA)算法的粗配准和基于K维树改进的迭代最近点(KD-ICP)算法的精配准相结合的点云配准方式，准确、高效地计算了扭角数值，为旋翼桨叶扭角的计算提供了一种通用有效的计算方法。最后，以某类型旋翼桨叶为检测对象进行了实验，结果表明扭角检测最大误差为54″，重复精度为18″，验证了系统的有效性和准确性。

关键词: 扭角, 旋翼桨叶, 数字化检测, 激光扫描, 点云配准