A Quantitative Evaluation Method of Aero-engine Blade Defects Based on Ultrasonic C-Scan

Abstract

Given the unintuitive and nonquantitative analysis of traditional artificial detection results of aero-engine blades, a characterization evaluation technique has been carried out based on ultrasonic defect point cloud. 3-D reconstruction display model of ultrasonic point cloud was completed by ultrasonic C-Scan imaging principle, spatial surface display technology, and 3-D visualization technology. A hybrid reconstruction method based on Delaunay triangulation and mesh growth algorithm was proposed, which achieved the 3-D reconstruction of a defect point cloud of regional discretization, the obtained results exhibited the spatial distribution of defect area, which intuitively provided the comprehensive view of 3-D defect model. Furthermore, the defect area was calculated via the vertex coordinates of each triangle in the reconstruction mesh. Taking flat-bottom holes as an example, an experimental model for quantitative evaluation was established based on the method. The quantization error of defect size is less than ±10% with comparison of the prefabricated defect size, so the feasibility and accuracy of the method are verified, which provides a convenient and effective way to quantitatively characterize the defects of curved components with immersion ultrasonic testing.