| As a key component of an aero-engine,the blade’s profile accuracy and surface quality are directly related to the engine’s operating performance and service time.Because the blade long-term in high temperature,high pressure and other complex operating conditions,the surface after work is very easy to produce scratches,scrapes,pits,group points and other small repairable damage,and then cause blade failure,resulting in engine failure.Therefore,regular inspection and maintenance of the engine and its key components are required both at home and abroad.Currently,manual polishing is still the main method for minor repairable damage to the blade surface,resulting in low grinding efficiency and poor overall consistency of the repaired surface,which is not conducive to mass production.Therefore,it is urgent to carry out research on key technologies for blade damage polishing and repair.To this end,this paper proposes a vision-based local CNC abrasive belt regrinding method for compressor blades to achieve efficient,high-quality and stable batch regrinding production of blades.The following research work has been carried out around blade damage area identification and positioning as well as CNC grinding methods:(1)The overall scheme of CNC abrasive belt local regrinding for blade damage area is designed.The structural characteristics of the compressor blade and the difficulties of local processing were analyzed,and the blade local grinding process scheme was developed.At the same time,the positioning method of the blade damage area and the CNC belt grinding processing method were initially determined.(2)The overall identification and localization method of the damaged area of the blade is determined.In view of the blade reflection phenomenon,the image processing process and classification method are proposed for the damage area markers,and the positioning of the damage area in the two-dimensional image is completed.On the basis of obtaining the three-dimensional point cloud of blade damage area markers,the improved point cloud alignment method is used to quickly and accurately obtain the corresponding points of the damage area marker data points in the theoretical model and solve the parameter coordinates and range of the damage area for the positioning problem of the measurement data in the CNC belt grinder.(3)Studied the target profile design and CNC grinding method for blade damage areas.The theory of curve and surface reconstruction was applied to optimize the initial target profile for determining the grinding level,and a calculation model for machining allowance in the damaged area was established.Through collision detection,the feasible range of the tool was planned,and the grinding path was modified and planned based on the expansion range of the grinding area.A material removal model based on empirical formulas and neural networks was established through experiments,and based on this,a process control method for precise removal of uneven materials in damaged areas was developed.(4)Developed blade defect identification and processing software.The software system was developed using C++and applied to the experimental platform to carry out damage area identification and CNC polishing experiments on a certain type of blade.The experimental results indicate that the method studied in this article can effectively ensure the accurate identification of blade damage areas and the quantitative removal of materials,and significantly improve the efficiency and quality stability of mass production of blade polishing repair. |
PDF Full Text Request