Abstract
In a design for manufacturing (DFM) approach the designer has to consider the interactions between the various parameters in the design and the ease with which it can be manufactured, very early in the design process. This research is aimed at providing designers with a methodology for early, quantitative evaluation of manufacturability. A hierarchical, domain independent model of manufacturability is first presented. Manufacturability is decomposed into five main areas namely, Compatibility, Complexity, Quality, Efficiency, and Coupling. The evaluation process begins with an examination of these issues as they relate to the design being evaluated. Next, the idea of manufacturability indices is introduced and contrasted with traditional measures such as cost. Examples of generalized manufacturability indices are provided. A generalized evaluation methodology which merges the hierarchical model with the idea of manufacturability indices is then explored. Two methods for comparing designs are discussed. The first method is based on a relative reference scale and makes use of priority theory for the evaluation. The second method utilizes a fuzzy linguistic approach to transform the index value into a goodness value. A hierarchical additive weighting technique is then used to combine the relative weights of the attributes in the hierarchy with the evaluations for the different designs. Several metrics for injection molded parts are developed to illustrate the application of the methodology to a specific domain. The ability to integrate expert knowledge into the evaluation makes the methodology particularly attractive to novice designers. Furthermore, its generalized nature ensures that the methodology is applicable to a wide range of manufacturing processes.
Shankar, Srinivasa Ravi (1993). A generalized methodology for manufacturability evaluation. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -1486811.