Modeling and optimization of quality and productivity for machining systems with different configurations

Evaluating and optimizing system performance in terms of quality and productivity is a challenging task for the design of both traditional and reconfigurable manufacturing systems. A set of new methodologies have been developed in this research to model quality performance for machining systems, the relationship between the system configuration and product quality, and between quality and productivity.; This research first presents a methodology to predict the product geometric quality for machining systems with different configurations. The methodology integrates two significant variation sources (i.e., kinematic and static variations) by modeling the system level variation propagation using methods such as Homogeneous Matrix Transformation (HTM), Finite Element Methods (FEM), Monte Carlo Simulation and Object-Oriented techniques. Experiments have been carried out to verify the proposed methodology.; This research then presents a methodology to model the fixture variation, which is a main contributor to product geometric variation. Two types of fixturing schemes, i.e., the 3-2-1 and the 4-2-1 schemes, have been discussed. The 3-2-1 scheme has six locators constraining the six degrees of freedom of a workpiece and is modeled by using the HTM and FEM, given the boundary conditions. The 4-2-1 scheme has one extra locator and is a statically indeterminate system. Its contact condition is determined using the Principle of Minimum Potential Energy and then used for calculating the variation propagation effect.; This research finally presents a methodology to analyze the quantitative relationship between quality and productivity for performance optimization. The relationship is modeled based on the analytical productivity model and the surrogate models for quality, which are generated from the modeling methodology of quality using the Design and Analysis of Computer Experiments. The optimal performance is achieved by maximizing the profit, which is a function of quality, productivity and cost. The impact of system configuration on quality is analyzed by using the surrogate models.; This research initiates a set of system level methodologies including the quality prediction and performance optimization. The methodologies are beneficial at both design and production stages of machining systems. The methodologies can be extended to other manufacturing fields such as assembly and diagnosis.