| Along with the development of modern manufacturing industry, many requirementsfor products in function and appearance have be proposed which makes the products’surfaces reveal two trends: complexity and diversity. The rise of the product surface’scomplexity brings numbers of problems to manufacturing process, such as how tocoordinate the relationship among surface quality, machining efficiency and machiningenergy consumption to meet the requirements of processing. CNC engraving andmilling is considered as a typical CNC machining way for complex surface with thefeatures of high speed, fast feed, small cutting depth and multi-axial processing, etc.CNC engraving and milling can well meet the requirements of complex surface fineprocess and becomes a most important machining way of complex surface in modernmanufacturing factory. Therefore, two critical problems of complex surface CNCengraving and milling are studied in this article.Initially, from the process characteristics，CNC engraving and milling is adequatefor finishing the complicated surface. However, a variety of complex surface shapesalways lead to these problems that the selections of cutting parameters, cutting tools andtool path patterns are unreasonable in CNC engraving and milling process planning. Itwill contribute to a solution of these problems if the shapes of complex surfaces areanalyzed quantitatively and evaluated accurately. As the complexity of surface in CNCengraving and milling is difficult to analyze quantitatively and evaluated accurately, adefinition and a calculation method of surface machining complexity (SMC) based onregional division are presented. According to the characteristics of CNC engraving andmilling and the product complex surface, seven typical categories of the geometryfeatures of surface machining and a definition of surface machining complexity areintroduced; Then, Three processes of surface regional division are presented, namely thesurface discretization based on isoparameter line sampling, coarse regions divisionbased on surface curvature and regions grouping based on criteria; After that, basedupon surface curvature variation, a calculating method of surface machining complexityis introduced in detail.Subsequently, the selection of cutting parameters is an important content in thecomplex curved surface CNC engraving and milling process planning. Whether theselection of cutting parameters is reasonable or not directly affects the performance indicators of CNC engraving and milling such as processing efficiency and quality aswell as the energy consumption. Therefore, this paper stresses concentration on theresearch of cutting parameter selection method on the basis of the surface machiningcomplexity. As a consequence, based on BP neural network, the cutting parameterprediction model of CNC engraving and milling is established. Combined with theevaluation method of weighted summation, this prediction model can be used to selectan optimized cutting plan. Aiming at establishing a mathematical model and BP neuralnetwork model to optimize processing time, energy consumption and the surfaceroughness, let the SMC and four cutting parameters as the cutting control variables.Later on, use the improved BP algorithm to train and test BP neural network model.Finally control the error within the allowable range, which is to say that the BP neuralnetwork model can be used for cutting parameter prediction in machining environment.Beyond that, the neural network module of Matlab is used for structure design, trainingand testing of BP network.Finally, a CNC engraving and milling case of relief handicraft is used to verify thecorrectness and practicability of the calculation method of surface machiningcomplexity and the selection method of cutting parameters in this paper. These methodscan provide useful guidance for engineers about how to select reasonable cuttingparameters of CNC engraving and milling. |
PDF Full Text Request