Case Studies

Computer aided design of materials processing is common now. Beyond one step forming operations, also whole manufacturing chains can be simulated. Conventional optimization problems for the manufacturing cycles are based on simulations of various variants of several processes according to the applied optimization technique. In the industrial problems finite element (FE) simulations are usually used for calculations of the objective function, which usually consists of dimensional accuracy of products and their in use properties, as well as tool life. Thus, solution of the optimization task is costly and there is a continuous search for alternative methods. The generic process models were proposed by Behrens et al. Application of metamodeling is possible, as well (see Section 2.1.2 of this book). These solutions improved the efficiency of optimization but the costs are still high. A new approach to solve optimization problem was proposed based on dynamic analysis of behavior in a group of coexisting agents. Review of the cited publications shows that various modeling techniques can be used, from advanced multiscale numerical models to the artificial neural networks. It is pointed out in the introduction to this book that selection of an adequate model to a particular simulation task is crucial for the efficiency of the optimization procedure. Thus, the selected examples of the case studies showing search for a compromise between predictive capabilities and computing costs of materials processing models are presented in the following sections of the book. The manufacturing chains for automotive part, pearlitic steel rails, and fasteners were considered.