刀(考古)
点云
涡轮叶片
计算机科学
点(几何)
准确度和精密度
汽轮机
功率(物理)
计量系统
涡轮机
维数(图论)
机械工程
工程类
数学
物理
统计
天文
纯数学
量子力学
几何学
作者
Junhui Huang,Miaowei Qi,Zhao Wang,Zijun Li,Jianmin Gao
摘要
The large-size blade is the core component of nuclear power and ultra-supercritical steam turbine, and its existing dimension detection methods generally have the problems of low detection efficiency or insufficient accuracy. So, an efficient and high-precision measurement technology of full profile based on multi view structured light 3D measurement is proposed in this paper. And several key factors affecting the measurement accuracy are analyzed. Then a highlight elimination technology combining multi view and phase shift is proposed to eliminate the effect of specular on the smooth metal surface of the blade. A non-overlapping point cloud registration method based on feature constraints is presented to solve the difficulty caused by the lack of overlapping point cloud on the thin inlet and outlet edges of blade and other areas. Finally, a platform for the whole profile of large-size turbine blade is constructed to realize the full profile measurement and evaluation of large-size turbine blades. The measurement results show that the full surface point cloud of an about 900mm long blade is completely obtained by the measurement platform, and the measurement accuracy of 70μm is achieved through the comparison test of high-precision gauge blocks and some standard balls.
科研通智能强力驱动
Strongly Powered by AbleSci AI