Aspects of decision making for metal powder production

Abstract

Production in the netal powder making industry is to a great extent a 'matching' procedure. Many base powders can be produred via the processes, but they will not match customer requirenents unless subjected to extensive 'manipulation'. As a result several 'control' stages are available. At production, at separation stage and at blending stage. It appears initially that complete control is possible. However, lack of reliable information on process capibilities and poor definition of products together with a complex blending problem make decisions extremely difficult. The work described in this thesis presents, it is believed for the first time, a coherent approach to rranagenent decision making in this industry. A multi-decision stage optimisation is suggested and case studies are presented. A variety of advanced and standard mathematical modelling approaches are utilised, including Dynamic Programming, Linear Programming, Simulation and Industrial Dynamics. The research concludes with an outline of further work required to productionise the recommended course of action.