Section Profile Measurement Of Aircraft Engine Blades Based On Chromatic Confocal Sensor

Blade is the most critical part of aero-engine,which plays the role of air flow guidance and energy conversion.The deviation of its shape from the design model has a significant impact on the energy conversion efficiency of the engine.Therefore,blade processing must meet strict tolerance requirements,and the quality of blade processing must be evaluated by measuring.Because the blade body is a complex free-form surface,its inspection can be transformed into the measurement and evaluation of the section profile.However,the existing methods can not take the accuracy,efficiency and integrity into account effectively.To solve the above problems,this paper designs an aeronautical blade section profiles measurement system based on the chromatic confocal displacement sensor.(1)Aiming at the problem of point cloud data acquisition of blade section profile,a 4-DOF blade profile measurement platform was constructed based on coordinate measuring machine(CMM),combined with precision turntable and non-contact chromatic confocal displacement sensor.According to the structure of blade tenon and CMM,blade fixture and sensor fixture are designed respectively,and multi-angle measurement is carried out by using zonal scanning method to obtain point cloud data of multi-section profile of blade body.(2)Aiming at the different coordinate systems of multi-view point clouds,the coordinate system is unified by using the rotation method.Firstly,a method to calibrate the position and pose of the sensor is proposed and solved by the mixed penalty function method.Secondly,the rotating axis and the workpiece coordinate system are calibrated.Finally,the three-dimensional measurement principle of the system is analyzed and the coordinate transformation model is established to unify the multi-view point cloud into the workpiece coordinate system.(3)Because of the low positioning accuracy of the turntable,the point cloud bifurcation occurs when the multi-view point cloud is converted to the same coordinate system by the rotation method.To solve this problem,a further and accurate method for integration based on overlapping point cloud registration and iterative closest point algorithm is proposed.Firstly,denoising is carried out based on curve checking method and statistical filtering algorithm;Secondly,NURBS curve fitting is used to increase sampling;Thirdly,overlapping parts of two adjacent point clouds are calculated iteratively and registered by ICP algorithm to obtain transformation matrix,and then the precise integration of multi-view point clouds is realized;Finally,the polar angle method is used to achieve the ranking of cross-section point clouds.(4)To solve the problem of point cloud segmentation for the leading and trailing edge and evaluation of characteristic parameters,a method based on curvature feature and RANSAC principle and parameter extraction algorithm are proposed respectively.In this method,the point clouds are roughly separated according to the curvature,and then the point clouds are accurately separated according to the principle of RANSAC.Then the chord parameters are extracted,and the segmentation for basin and back is completed by using the chord equation.Finally,the arc is extracted by the method of equal radius,and the maximum thickness is calculated.The system based on measurement scheme is built,the calibration and aeronautical blade measurement experiments are carried out.The results are compared with those of CMM.The experiment shows that the indication error of the measurement system is less than 15?m.