Analysis And Research On Multi - Source Error Model Of Machine Feeding System

Manufacturing is an important basic industry to provide technical equipment for the various sectors of the national economy and national defense construction.The CNC machine tools has been the focus of research as the basis of advanced manufacturing technology equipment.Precision,as an important indicator of performance of CNC machine tools, has long been the most important emphasis in the field of machine tools research.Feed system is a key component which affect the accuracy of CNC machine tools.Its quality and performance have a significant impact on CNC machining precision.Feed system is the research object of this dissertation. This dissertation focus on error mapping, the error modeling of feed systems component layer, optimizing of machine tool volume error model.About error mapping. Aiming at the shortage of the 9-line method based on laser measurement used in identification of geometric errors of CNC machine tools, in order to improve the accuracy of the error identification, by analyzing the principle error of 9-line method, a new method of geometric error identification on the basis of 9-line method was proposed. The method correct the errors in straightness measurement and optimize the selection of measurement point and coordinate position to find the preferable result of the error identification. The method was applied on a linear feed system, and the result showed the feasibility and reliability of the proposed method.About component layer error modeling of feed systems,by considering the affect of multiple parts and multiple parts to the feed system,an innovative coupling modeling method is proposed to show the mechanism how the components errors,assembly errors and contact stiffness of components affect the accuracy of feed system.About optimizing of machine tool volume error model,for lack of multi-body system theory modeling method,combined withthe mechanism of componentaffecting the error of the feed system,a optimizing method based on non-rigid error matrix modeling is proposed.