Research On 3D Visualization Nondestructive Testing Technology For The Aircraft Engine Blade

In recent years,with the rapid development of national aviation field and the improvement of aviation industry's quality standards for aviation components,the software platform of ultrasonic three-dimensional visualization nondestructive testing has developed rapidly,and the technology of ultrasonic three-dimensional visualization has gradually extended from the medical field to the industrial field.In the industrial field,the research on ultrasonic nondestructive testing of aero-engine blades has always been carried out around improving detection efficiency and reliability.This paper aims to develop an ultrasonic three-dimensional visualization non-destructive testing software system.The blade detection efficiency and reliability will be improved by studying the key techniques of the reverse measurement process and the automatic extraction technology of the boundary feature points of the defect area.The main contents are as follows:(1)Aiming at the current situation that the ultrasonic three-dimensional visualized non-destructive testing system has poor expandability,low generality and long development cycle,by analyzing the detection principle of ultrasonic C-scan detection and combining with the actual needs of the project,a UML and component technology is proposed to rapidly develop a universal modeling method for ultrasonic three-dimensional visualization non-destructive testing system.The ultrasonic three-dimensional visualization non-destructive testing system model is constructed,including the architecture,requirements,use cases,static classes,activities,timing and component deployment models.Finally,the software system prototype is developed under the C#.NET platform,and the effectiveness of the modeling method is verified by the measured data.(2)In order to quickly obtain the specified points of the three-dimensional point cloud model,and facilitate the manual selection of the target points of the three-dimensional model in the reverse measurement process,the picking method of the point cloud model is studied.The built-in picking mechanism of OpenGL is analyzed.Aiming at the insufficiency of the picking mechanism applied to the ultrasonic three-dimensional visualization system,a ray-based picking method is adopted in the project software.At the same time,combined with the geometric principle of computer graphics,the intersection algorithm between ray and triangle mesh is realized by calling OpenGL graphics library development technology using C#.NET,and the transplantation of the algorithm is completed.The algorithm effectively processes the intersection process of the triangular surface patch and the ray after the pointcloud data is triangle meshed,thus providing a complete and effective picking scheme for the ultrasonic three-dimensional visualized non-destructive testing system.(3)In order to accurately and automatically extract the feature boundary points of the blade ultrasonic C-scan cloud model,the algorithm principle of the mean drift clustering algorithm is studied.Taking the spatial position of the point cloud as the clustering feature space,the boundary feature point extraction of the leaf point cloud model is realized.Combined with the data type and characteristics of ultrasonic C-scan data,a multi-scale mean shift algorithm based on spatial position and ultrasonic intensity value is proposed,which realizes the automatic extraction of boundary feature points of faulted blade defect areas.