UBC Theses and Dissertations
Dynamically reconfigurable machining systems Oldknow, Kevin David
The thesis discusses the use of Dynamically Reconfigurable Machining Systems (DRMS) as a means for manufacturers to enhance their machining capabilities on an as-needed basis. This is motivated by a market characterized by decreasing product life cycles and accelerating rates of change, requiring manufacturers to acquire a broader range of capabilities for shorter durations and challenging the economic feasibility of the permanent acquisition and utilization of machining systems that cannot be rapidly and automatically reconfigured. The successful implementation and utilization of DRMS requires a system architecture that allows physical hardware and software components to be automatically integrated into the system, and utilized effectively in practical cutting operations. The primary objectives of the work presented in this thesis are to provide technical advances that both improve the efficiency of existing control systems and illustrate new approaches to the problem of dynamic reconfiguration. Particular focus is placed on the improvement of both performance and accuracy of machining operations. A novel control strategy, referred to as three-dimensional dynamic interpolation (3D²I) is presented as a means to facilitate both system dynamics and process compensation through manipulation of the reference trajectory in tangential and orthogonal (axial, radial) directions. The mechanism is implemented within a novel Field Programmable Gate Array based architecture as a means for validation. Results from both simulation and experimental work investigating the use of the 3D²I mechanism for contour error reduction during high-speed contouring are presented, demonstrating the utility of the approach with particular emphasis on cases in which nonlinearities such as amplifier saturations occur. It addition, a novel and efficient means for on-line estimation of contour error is presented. The utility of the 3D²I mechanism for process compensation is shown through the implementation of a multiple constraint based tool deflection compensation strategy, in which orthogonal offsets to the reference trajectory are used to compensate for finished surface error due to tool/spindle deflection. A novel approach is used, in which average and low frequency force signals are used to identify process parameters. These parameters are used as inputs to a validated model in order to estimate and correct for surface error. An automated approach to the integration of process monitoring utilities is presented, in which a distributed process variable mechanism is used to propagate the required information to process monitoring resources that have been integrated within the system. This allows arbitrary modules to play an active role in the machining process. Finally, a formulation of machining economics that incorporates the stochastic nature of machining parameters is developed. This formulation allows the economic impact of process monitoring resources to be assessed, and provides a basis for decision making in process monitoring resource allocation.
Item Citations and Data