Research On Identification Model And Method For Design Defects Of Machanical Products & Its Application In Mechanical Transmission System

In the design process of mechanical products, verification work is an important task. Since function of verification system is incomplete, design defects of mechanical products are very difficult to be identified in time. Therefore, the efficiency of design and development of mechanical products are decreased, and design costs of mechanical products are also increased. Based on constraint model between design parameters of parts, design defects of mechanical products can be found quickly and verified accurately by using design defects identification model. The design quality and development efficiency of mechanical products are improved effectively by applying the identification model. In this paper, focusing on mechanical transmission systems, the constraint model is established, design defects identification model and its modification methods for strength design defects, assembly design defects and machinability design defects are investigated intensively and systematically.1. According to constraint model between driving and driven gear of non-circular spur gear, design defects identification model and modification method of non-circular spur gear are proposed. Design defects such as discontinuous point and cusp are analyzed, identification method of design defects with discontinuous point is given. The pitch surface cusp of high-order arithmetic spiral or quadratic curve bevel gear is modified by using pitch surface of cutter to replace pitch surface of non-circular bevel gear in cusp.2. A design defects identification method of mechanical products based on regress analysis model is presented. Taking rzeppa constant velocity joint(RCVJ) as the research object, the contact stress between inside and outside channel of RCVJ is treated as a basis for determining strength of RCVJ, the historical design parameters for the previous design schemes are treated as dependent variables, the regression equations between contact stress and its influence factors are established, by analyzing regression equation, the influences of various factors on the strength of inside and outside channel are derived, the maximum fitting errors for various influence factors are also calculated from regression equations. Design parameters of new design schemes are introduced into regression equations then, the differences between the regression equation outputs and the design parameters are computed, by contrasting with the maximum fitting errors, the design defects on strength of RCVJ are identified.3. A assembly design defect identification method of mechanical products based on verification tree model is proposed. Aiming at problems of assembly constrain verification in mechanical products design, a data structure of verification tree is adopted to store information of assembly constrain relationships into node of tree and to store modified design parameters into the verification modifier, thus the changes of contents in verification modifier are monitored by monitor, and the start-stop of verification process is decided. Therefore, a prototype system of dynamic relating assembly verification is developed to verify design defects of driving leg assembly.4. Identification model and modification method of mechanical products design defects are investigated based on constraint relationship net. Association method for assembly design parameters of mechanical products and its constraint relationship judgement functions is proposed, data structure for nodes of assembly constraint relationship net is studied, construction method for assembly constraint relationship net is given, taboo list travel method for assembly constraint relationship net is descripted, identification models and methodology for assembly design defects based on constraint relationship net are proposed. Starting from design defect node of assembly constraint relationship net, taking the sum of constraint path weights as object function, correction paths for assembly design defects are optimized utilizing ant colony algorithm. Finally, design defects of diaphragm clutch are identified and modified by using above method.5. A design defect identification method of mechanical products based on artificial immunity principle is presented. Geometric constraints relationships between mechanical parts are formulated as design defect discriminat functions, and assembly constraint relationship matrix is established based on encoding of geometric constraints between mechanical parts, in this matrix, directed graph is utilized to store information of matrix, the node of directed graph is used to identify parts number, the design parameters of mechanical parts is stored in the data structure of node, and the weight of directed graph arcs is treated as antibody and design defects of mechanical parts treated as antigen, mution rules antibodies are determined according to the affinity between antibody gene and corresponding antibody gene. Through mutation operations for antibodies such as trimming, splicing and replacement, the dynamic recognition of mechanical products design defects is realized based on decoding analysis for weights of directed graph arcs.6. Simulation verification identification methodologies for design defects of high-order non-circular gear pairs are presented, these design defects include undercutting, overcutting and pitch curve interference. A precise mathematical model of generation cutter for non-circular gear pairs is established. Based on meshing relationship between non-circular gear and its generation cutter as well as meshing relationship between driving and driven gears, generation algorithm of teeth profiles and meshing simulaition algorithm of driving and driven gear for non-circular gear are proposed. Simulation verification model of high-order non-circular gear pair is established by using above algorithm. In addition, design defects modification methods for undercutting and overcutting of high-order non-circular gear pair are presented.