Patterning algorithms for operation clustering for reconfigurable machining systems

One of the basic characteristics of a Reconfigurable Machining System (RmS) is the Customized Flexibility. That means that machines and systems are designed with a partial flexibility, that although limited compared to CNC machines, it allows to machine all the parts of the family. This may bring substantial economic benefits by reducing the machine cost and enhancing productivity. To enhance productivity, features in a single part have to be grouped together and machined simultaneously using a spindle head with parallel tools. To reduce cost, it is worthwhile to check whether the same spindle-head may be utilized a few times on the same part, or the entire part family. The first step is, therefore, developing a methodology that allows an automatic grouping of identical feature-clusters for a single part as well as for the whole part family.; Previous research in operation clustering is limited to the CNC domain only. The grouping principles strongly depend on the assumed single cutting tool configuration. But one of the principles of RmS is to adopt the customized designed multi-tool simultaneous cutting (i.e., parallel machining) configuration that enables to achieve high productivity. Therefore, the existing research results are inadequate in the RmS domain.; This thesis develops a new methodology that incorporates a set of novel hole pattern identification algorithms for operation clustering This set of algorithms generates common parallelism-based operation clusters to maximize the usage of identical spindles or machines. Using translational and rotational vectors, the algorithm effectively solves the challenging task of automatically extracting common operation clusters. Using the relationship between translational vectors and common operations from a single part, the algorithm successfully identifies the features that may be combined for a repeated single operation. Using the relationship between translational vectors and common operations from all member parts in a part family, the algorithm effectively extracts the common and volatile operations. The extraction of common and volatile operations at the part family level enables a system designer to determine the reconfiguration capabilities of the machines. Further using the relationship between translational vectors and tool arrangement options, the algorithm generates optional solutions. These optional solutions enable a system designer to improve the system balance and design for scalability, to accommodate a part family, and to reduce the overall system cost.; This research may benefit also the Design for Manufacturability. An extension of the methodology will point out which features are candidates for location modification, which consequently would reduce the total machining time.; In summary, the methodologies developed in this research provide an original and effective way to perform operation clustering in the RmS domain. The issues addressed are an important part of the RmS science.