Key Techniques For Mirror Milling Of Large Aircraft Skins

The aircraft skin consistutes its profile whose quality dictates the aerodynamic performance,while the processing rate of the aircraft skin directly affects the entire productivity of the aircraft.In order to gain higher performance,the new generation of domestic large aircraft is designed with more advanced aerodynamic layout,demanding larger but more complex skins,as well as higher manufacturing accuracy.Mirror milling has been nominated as a trending technology for machining large skin parts featured with thin wall and complex profile,for its capability of achieving constant thickness of the skin part by synchronizing the motion of the tool and its mirrored support.Existing mirror milling process is incapable of achieving such high accuracy and efficiency which in particular is caused by the intricate process constraints for the tool path generation and the shape unstability of the raw skin sheet.This thesis has conducted a thorough study focusing on the mirror milling process including the tool path generation,skin surface inspection and the tool path transformation.The main contribution of this thesis is listed below:(1)In order to generate eligible mirror milling tool path for complex skin parts,the desired tool path should satisfy geometric constraints of the mirror milling process including the stepover distance constraint,crossover constraint,tool retraction constraint,etc.We proposed to deal with this issue by establishing an image representation for the machining feature as well as for the tool path itself.Based on this,a tool path generation and optimization scheme is proposed by first propagating initial path from the extracted medial curve,and then being treated as binary image and iteratively deformed back into the target region.The noval image-driven scheme turns a global tool path optimization issue into image operation process,and gained over 20% increase of machining efficiency without violating the process constraints of the mirror milling,as compared with existing tool path generation algorithm.(2)In order to inspect the in-situ skin surface,whose actual shape largely deviates from the design model,we proposed multi-source integrated fusion(MSIF)method for the inspection task.First,the correlation between high-accuracy data using touching probe and low-accuracy data using laser scanning were established,which helped to analyze the effect of system error and random error towards the fusion error.Based on this,data fusion error was reduced using weighted residual approximation as well as using multi-source stacking method,which facilitates the fusion task for data with different level of accuracies.The proposed MSIF increases the inspecting rate by over 2.5 times compared with existing methods using single source,while maintaining desired accuracy.(3)In order to efficienctly and accurately transform the nominal tool path onto the actual measured surface,an adaptive tool path transplantation method was studied.The error distribution was defined over the skin surface according to the inspection results.Based on the error distribution,a surface segmentation method was proposed to divide the surface into three regions,while each region together with the transitional one was assigned with a particular transformation strategy to adaptively accommodate the nominal tool path with the actual surface.Compared with existing tool path transformation method using single strategy,the proposed method achieved qualified accuracy for both the thickness(thickness error within±0.1mm)and the profile of the skin,and over 10 times increase of efficiency against recomputing of the tool path.(4)Based on the above studies,an offline numerical control programming system and an adaptive machining system for skin mirror milling were developed.Results successfully applied to the research and development of the C919 skin parts.Preliminery experimental results of the testing skin parts indicate qualified accuracy of the C919 requirements,and over 2 times increase of entire machining efficiency,including 4 times increase of NC programming efficiency,2 times increase of skin inspection rate,20 times increase of tool path transplantation and 20% increase of machining efficiency.