Digital Measurement And Tolerance Evaluation Of The Large Stator Blade

The pursuit of product functionality and appearance become increasingly strict inthe industrial fields which is closely related to manufacturing industry such as aviation,spaceflight, shipbuilding, automobile and mold. The contour design、the processingquality and the detection precision of free-form surface product are proposed morestrict requirements. The digital measurement and error evaluation technology of thelarge complex curved surface parts have been used widespreadly during themanufacturing process. Meanwhile, the digital measurement meets with the modernmanufacture’s requirement of integration with CAD/CAM/CAE. Therefore,researching on the digital measurement and error evaluation technology of largecomplex curved surface parts is significant for improving the product quality andproduction efficiency. The digital measurement and error evaluation of large statorblade are researched based on the digital measurement, data processing, datummatching and error evaluation.The various digital measurement methods were discussed for the structure oflarge stator blade. The3d laser scanning measurement method was researched deeplyand compared with CMM. The measuring methed of large stator blade was established.The principle based on moving station of3d laser scanning measurement was analyzed.The point cloud data of the large stator blade were acquired through taking five spatialdatum points.Simplification and removal of noise points were researched. The sources of noisepoints were analyzed. Algorithm based on octree of fast k-nearest neighbor wasproposed to remove the noise points of the large stator blade point cloud. Thedenoising results showed that the algorithm had a favorable characteristics. Theadvantages and disadvantages of various kinds of data reduction methods wereanalyzed and compared. A mixed algorithm that combines bounding box with curvature sample method was proposed to reduce the measurement data of large statorblade and gained certain effect.The datum matching algorithm were investigated for matching the point cloudand the theoretical CAD model. Some improved ICP algorithms were evaluated andcompared. A multi-step algorithm was proposed for the datum matching of large statorblade, namely adopted the algorithm of calculating curvature to achieve initialmatching, then using ICP-SVD algorithm to reach precise matching. The correctnessand effectiveness of the datum algorithm were verified which laid a foundation fortolerance evaluation and its effectiveness. Then an approximation algorithm for point-to-surface distance was adopted. The error results of large stator blade were obtained.Finally tolerance evaluation of large stator blade was achieved and analyzed.Through the research on the digital measurement and tolerance evaluation of thelarge stator blade, the error distribution and deformation of large stator blade wereobtained. Research results can be provided references for the optimal designs of bladedie casting and the following machining. Research methods can be applied thefollowing heat treatment and the machining process for the large stator bladeinspection.