Design and synthesis of metal forming processes

In this globally competitive environment it is imperative to devise new manufacturing methods that are cost effective and continually seek to improve the existing manufacturing techniques. Computer-aided-design (CAD), computer-aided-manufacturing (CAM), and computer-aided-engineering tools have been harnessed in the previous decade to analyze manufacturing processes. The availability of computing power has also increased dramatically which allows more computer intensive operations. However, most of the CAD/CAM/CAE tools are built around iterative "input-analysis-input" techniques. This methodology is cumbersome and time consuming, especially if the analysis problem is quite large.;The primary focus of this dissertation is on the development of design methodologies in metal forming and manufacturing. The methodologies presented are generic and may be applied to a number of processes. The implementation of numerous tools gives these methodologies their versatility and effectiveness. The selection of a methodology for a particular operation depends on the task at hand. The design methodologies presented include: (1) Hierarchial Design Methodology; (2) Intelligent Hybrid System; (3) Taguchi and Axiomatic Design Process; (4) Property Based Design.;This hierarchial design methodology is time and cost effective because it provides a systematic means of achieving an optimum process design by limiting the design space in successive, orderly steps. Conceptual design using the axiomatic design principle is used to identify important parameters. Preliminary designs consisting of simplified tools are employed to develop an initial design. As the design space is narrowed more accurate finite element methods are used to analyze the process.;The second approach further improves the hierarchial design concepts and is performed by artificial intelligence tools. The artificial intelligence approach to metal forming processes demonstrates the use of an intelligent hybrid tool to design and control the pack rolling process which is presented as an example. This intelligent harnesses the power of two AI tools--the expert system and the neural network approach.;A new experimental design methodology is presented which is particularly effective if the analytical solution is not available or the process is too complex to model. In this situation, carefully controlled experiments are employed to provide the understanding of the process and to establish a design criteria for control of the process. This methodology employs the axiomatic design approach combined with the Taguchi design method to generate a conceptual design. This methodology works very effectively with any experimental technique and has extremely high merits if used for design of a relatively new process.;A novel property based design approach is presented which will have tremendous impact on the metal forming community. The objective of this methodology is to design the process based on the desired properties of the final product. This methodology brings a higher level of scientific approach into the design process. The property based methodology is developed and demonstrated for the extrusion process. A new extrusion die design is presented based on controlling the Zener-Holloman parameter which yields a uniform microstructure along any section of the extrusion. Moreover, this approach reduces the scatter of the grain size along the cross-section and also achieves the desired crystallized grains at each section of the extrudate.