Process modelling and adaptive control of a metal milling process
Within the aircraft industry, a large number of the aluminium components are milled entirely from single solid billets of metal. Current practice is to machine at conservative and constant feed-rates. This avoids damage to the workpiece and therefore reduces costly scrap. On-board monitoring systems on milling machines within the aerospace industry have introduced the possibility of continuously and adaptively varying the feed-rate online. It is proposed that this variation could be based on cutting power. Machining parameters, such as depth and width of cut, and tool efficiency, can vary considerably during the manufacture of a single component. It is therefore likely that an adaptive controller should be used. NICS is an adaptive controller algorithm that has been under development since 1990 Stoten et al. It is known to have good properties with respect to robustness and controller performance and is therefore being investigated for use on the milling machines for aluminium components. Substructuring is a dynamic testing technique that combines numerical models and physical tests into a single hybrid experiment. Briefly, it allows the engineer to "glue" a real component (in a lab) and a virtual component (implemented on a computer) together. These real and virtual components can then be tested dynamically and in real-time as if they were a single, complete component. Hence, substructuring provides the engineer with a framework to allow testing of individual components from entire structures at full scale without requiring the complete structure to be physically present. Accurate, hi-fidelity control is essential to the successful reproduction of structures using this technique. Hence, the goal of the work presented in this thesis is the development and implementation of control strategies for use in the field of metal machining. A scheme that the controls the cutting force transmitted to aluminium workpieces by varying the feed-rate is developed. A requirement of this scheme is that it be simple to set up and implement whilst, at the same time, regulate the cutting power in the face of process para peter variations. The ability of the adaptive controller to meet these requirements is demonstrated both in simulation and on a pilot rig developed in the work. A model that relates peak cutting force to measured cutting power is developed for use with the aforementioned control scherne along with high-level control software implementing this scheme. Rirther, controllers both fixed-gain and adaptive, are developed and tailored for use in the substructuring technique. The performance of the developed controllers is investigated both in simulation and in the laboratory on a linear large-scale hydraulic system. The advantages of using the adaptive controller scheme over a fixed-gain linear equivalent are presented. An example demonstrating the use of the substructuring technique in the field of machining is discussed, including a specific application.