Process sequence design for cold forging is an experience-oriented task which depends on designer's skill and decision. Therefore, several expert systems for the design of cold forging process sequence have been developed in order to reduce the time and cost required at design stage. However, there is a number of designer's own rules which might conflict with general design rules, and specification of forming facilities has not been considered in the developed systems. Although the existing expert systems previously developed have made many progress by developing algorithm to automatically regenerate design knowledge and constructing the necessary database, there are difficulties in setting up the proper hierarchy in implementing proprietary know-hows to the existing design rules. Therefore, in the present investigation, searching techniques were introduced to an expert system in order to simulate the heuristic nature of cold forging process design.
Due to the non-deterministic nature of process sequence design for multi-stage cold forging, various process designs are available depending on the initial billet geometry and the order of basic processes such as forward/backward extrusions, upsetting and trimming process. Therefore, various process sequences should be determined and compared to obtain an optimal solution. For this purpose, $A^*$ and depth-first searching techniques used in artificial intelligence were introduced in developing expert system for multi-stage cold forging process design. However, it was found that depth-first search was more appropriate to cold forging process design than $A^*$ search since all feasible solutions can be presented to process designers and it is easier to compose an evaluation function for depth-first search than that for $A^*$ search.
As a result, various possible process sequences can be designed by the introduced depth-first search and an optimal solution can be searched by estimating the values of evaluation functions which are introduced to represent the important design characteristics. In this study, the distributions of global effective strains in the final product and forming loads required at each forging stage were selected to be controlled. In addition, a more realistic process sequence can be designed in the developed system by considering manufacturing conditions such as the number of forming stages, forming loads, shearing diameter of the coil, open upsetting diameter, and knockout lengths of the dies and punch. In this dissertation, a methodology of applying searching techniques for the process sequence design is discussed, and the flexibility of the introduced searching techniques is evaluated by generating design examples of a shaft part, a hexagonal and wrench bolt made of AISI 1045 and a clutch part made of AISI 1020. For verification of the developed expert system, finite element analyses and forming experiments of a cold forging process sequence of pin-shape part were also carried out. As a result, a pin part was successfully forged without defects based on the designed process sequence obtained from the developed expert system. In addition, in order to extend the possible design geometry, a process sequence design for a semi-axisymmetric part with asymmetric geometry which can be forged by axi-symmetric processes such as upsetting, forward and backward extrusion was conducted.
Ductile fracture is one of the most important design variables in cold forging since materials have forming limits. Therefore, In this dissertation, the limitation and applicability of ductile fracture criteria based on a work hypothesis and Cockcroft and Latham were investigated. For this purpose, experimental and numerical investigations of simple upsetting were conducted for aluminum alloys Al1100-O, Al2024-T3, Al6061-T4, and Al7075-T4. As a result, the fracture mode of each alloy was observed. The study was extended for pin-shape cold forging of Al1100-O and Al6061-T4 to compare the likeliness of fracturing according to two criteria. As a result, it was found that Cockcroft and Latham's criterion might be useful in cold forging process design.
