Laser Line Measurement And Data Processing Methods For Casting Turbine Blades

Aero turbine blades is subjected to high temperature,high pressure and heavy load environment and are usually formed by precision casting.Due to the uneven distribution of thermal stress,thin-walled structure and other factors,the bending and twisting deformation of the blade during the casting process is very common.The detection of the profile and the quality control are extremely important.The existing contact measurement method is inefficient and the data processing process is complicated.This paper takes the laser scanning and measuring equipment of the blade line as the object,studies the line laser scanning rotation platform calibration method,the laser stripe center line extraction method,and the blade section data processing,which mainly includes:(1)Using a laser tracker and a ceramic double ball plate,the coordinate transformation relationship between the local coordinate system of the laser tracker and the local coordinate system of the line laser sensor was established,and the calculation equation of the central axis of the rotating table using the spatial standard ball as a reference was derived.The coordinates of the central axis and the center point of the rotation platform were solved by least squares fitting method;the calibration experiment was designed and completed on the laboratory laser measurement platform,and the calibration ball diameter measurement error was controlled within 0.02 mm after calibration.(2)Analyze the phenomenon of uneven distribution of gray value in the central cross section of blade front and rear edge regions,establish a mathematical model of offset Gaussian distribution,and propose an offset-free subpixel centerline extraction algorithm.The subpixel edge point calculation equation is based on surface fitting method,use the minimum inscribed circle of the edge points as the constraint,the centerline coordinate is obtained by iterative calculation of the circle center;the fringe image is generated for simulation experiments,and the standard ceramic ball measurement experiment is performed on the laser measurement platform.The result shows that the pixel error is within 0.1 pixel,the error of the standard ball diameter is within 0.015 mm.(3)Complete point cloud data processing module in the iCloud3 D software environment,including point cloud data import,point cloud extraction,design model cross section extraction,single section point cloud sorting,and test report preservation;study automatic extraction of cross sections based on k-means clustering and outer-contour point cloud sorting based the concaves model,which achieves the conversion of the two-dimensional scattered point cloud to the ordered line model,and the output line model is applied to point cloud matching and error calculation.(4)Using a certain type of turbine blade sample as the object,complete multi-section point measurement and point cloud data processing of turbine blades,including cross-section extraction,noise removal,single-section point cloud sorting,point cloud matching,automatic error generation,report generation.The detection process is completed in less than 2 minutes.The use of batch processing can achieve automatic data processing for the same kind of turbine blades,and the efficiency of the existing manual treatment of the enterprise is improved by more than 70%.