Development of an expert system interface for the automated analysis of the sheet drawing metal forming process using finite element analysis techniques

Abstract

The objective of the work presented in this thesis was to develop an expert system for sheet metal forming. In doing so an examination, of implicit and explicit Finite Element Analysis (FE.A) codes and methodologies was conducted. Guidelines as to how best to carry out simulations of sheet drawing, were developed. Work was conducted into deep drawing and shallow drawing of non complex shapes, because of the many analytical, and empirical studies conducted on cup and box drawn shapes. The process parameters effecting drawing of complex parts are present in drawing of non complex parts such as cup and box shapes. A valuable insight into drawing of complex parts can be attained economically, with the proposed study of drawing of non complex parts, such as cup and box shapes. Modelling and solution variables present in implicit and explicit F.E.A. codes are explored by the modelling of non complex parts such as cup and box shapes. Deep drawing is a common process used in the manufacture of auto body components; cans, cups, bathtubs, sinks, and other similar items by drawing rolled sheets into there final geometric form. Configuration of a new deep drawing process is highly empirical with many parameters determined by trial and error. Large deformation numerical Finite Element Analysis of the deep drawing process can be applied to shorten the process reconfiguration time and remove the expense of trial production runs of parts. In the research work reported in this thesis parametric studies are performed on critical deep drawing parameters. The findings are presented and are incorporated into the Expert System for Metal forming. Material non linearity is introduced by a description of drawing blank material plasticity. The discussion on plasticity introduces rate independent material behaviour models, typically used to model the behaviour of Steel and Aluminium Blanks. The Taguchi method of experiment design and results analysis is applied to develop strategies for the design of best practice finite element simulation of cup drawing. The expert system for cup drawing contain domain knowledge in the form of analytical model and dedicated automatic development of Finite Element Analysis simulations which is developed in ANSYS command script files. The control aspects of the expert system utilises goal and data driven reasoning within the control algorithm. A blackboard architecture to the expert system is promoted with the ANSYS Advanced Parametric Design Language (APDL) macro programming language.