A Fuzzy Set AHP-based DFM Tool Under Concurrent Engineering Environment

In the light of growing global competition, organizations around the world today are constantly under pressure to produce high-quality products at an economical price. The integration of design and manufacturing activities into one common engineering effort has been recognized as a key strategy for survival and growth. Design for manufacturability (DFM) is an approach to design that fosters the simultaneous involvement of product design and process design. The implementation of the DFM approach requires the collaboration of both the design and manufacturing functions within an organization. At present, For some reasons DFM approach is ineffectively including lack of interdisciplinary expertise of designers; inflexibility in organizational structure, which hinders interaction between design and manufacturing functions. Design for manufacture is the practice of designing products with manufacturing in mind. Early consideration of manufacturing issues can shorten product development cycle time, minimi overall development cost and ensure a smooth transition into production. In this paper, part manufacturability under Concurrent Engineering (CE) environment was analyzed in detail. An evaluation system of DFM was proposed according to CE ideas. A fuzzy set-based manufacturability evaluation algorithm is formulated to generate relative manufacturability indices to provide product designers with a better understanding of the relative ease or difficulty of machining the features in their designs. An analytic hierarchy process (AHP) method is introduced to assign weighting factors to features to reflect their functional importance. Results from the case studies show the method available and practicable. Introduction Design for manufacture (DFM) is the core idea and implement way of concurrent engineering, means considering the correlative factors at the same time of designing product. It is intent to realize the concurrent of product design and product manufacturing process design, consequently to improve design quality and quicken research schedule [1]. Recent years there are already several commercial manufacturability analysis software systems [2, 3] and various research and academic systems [4–8]. These systems differ greatly in their scope, measures, methods, and the levels of automation. However, most have (a) a limited scope: they either focus on material/process selection or perform detailed analysis for a specific manufacturing process; they either give re-design suggestions based on rules-of-thumb/analytical results or estimate the manufacturing time/cost, (b) the measure and logic used in these systems are decided by vendors/developers and the user cannot set a preferred measure, (c) the manufacturing knowledge is domain-specific and hard-coded. The knowledge is either hard-coded or stored in specific database in these systems. It is not possible to update or customize the system with the specific shop manufacturing capabilities. In effect, these systems imply a ‘universalization’ of manufacturing capabilities and costs even though individual shops vary widely. The major considerations for developing flexible and customizable DFM software are: process and knowledge representation, DFM measures, and software framework. It involves many research challenges such as manufacturing knowledge representation, retrieval and storage, manufacturing Applied Mechanics and Materials Vols. 10-12 (2008) pp 145-149 online at http://www.scientific.net © (2008) Trans Tech Publications, Switzerland Online available since 2007/Dec/06 All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of the publisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 130.203.133.33-14/04/08,12:12:31) process/resource characterization, geometric reasoning (feature recognition/attribute extraction), process planning, artificial intelligence (rule-checking), manufacturing process analysis and so forth. Most importantly, people do not have an overview of manufacturability analysis and there is little work completed on specifying the generic steps to perform manufacturability analysis tasks. The second issue arises mostly because there is not a consistent definition for ‘manufacturability analysis’. Shah and Wright [8] reviewed commonly used DFM measures and classified them into either qualitative or quantitative. Gupta et al. [6] grouped measures of manufacturability into binary measures, qualitative measures, abstract quantitative measures and manufacturing time and cost. The third issue requires manufacturability analysis systems to be customizable and expandable. Besides using a commercial software tool, two practices are often adopted to develop in-house manufacturability analysis applications. They are building tools based on a CAD system’s application programming interface (API) such as UGS APIs, and developing applications directly over solid modeling kernels such as ACIS or Parasolid. Since much manufacturability analysis applications have already been developed, future manufacturability analysis systems should be built with a fully open architecture. This allows easy integration of a broad range of existing manufacturability analysis tools, rather than discarding them and building a new one from the scratch. DFM Integrated Manufacturability Evaluation System Part manufacturability includes many contents: economic index and technology index, for instance. Machinability is the most direct content of DFM and the main content of influencing technology index, and the radical content of the cooperation between design and manufacturability in the same time. This article focuses on the machinability of manufacturability. In order to satisfy the functional requirements a DFM shell architecture is shown in Fig.1. The shell consists of two interfaces for the knowledge engineer and the designer. The data and knowledge base, and three evaluation modules. The knowledge engineer enters the domain knowledge into the shell through the knowledge setup interface. The designer interacts with evaluation modules through user interface to choose DFM measures, perform analysis tasks such as process selection, feature recognition, rule-checking, process planning, and estimation, and get advisories, feedback and estimation results. The designer can also access data and knowledge base through user interface if he/she wants to modify the settings. The data and knowledge base contains all the data and knowledge used in evaluation, such as manufacturing process and resource data, material data, cost/time estimation structure, good practice rules and part information model. To perform manufacturability analysis, 3D CAD model needs to be enriched to smarter information models such as a set of high-level attributes or the manufacturing feature model. Since manufacturability analysis cannot be done in one step, three generic steps are classified in the evaluation module: process pre-selection, qualitative analysis and quantitative analysis. Hence, different levels of abstraction for manufacturability analysis are supported. The three manufacturability evaluation modules interact with data and knowledge base at different levels. The process pre-selection module uses high-level product and process attributes. In this module, the manufacturing processes and the part are modeled using a set of attributes and no other detailed information has been involved yet. The qualitative analysis module applies heuristic rules, based on recognized features, part material and manufacturing process and resource information. The part is represented at the feature level and the manufacturing process capability is modeled mainly using geometric constraints at this level. The quantitative analysis module involves with all kinds of detailed information including manufacturing cost/time structure and formula, manufacturing operation details and first-cut process plans. Machinability Evaluation Machinability evaluation is hierarchical, includes qualitative evaluation and quantitative evaluation in general. The former only make an estimate of yes or no, namely estimate whether the designed 146 e-Engineering & Digital Enterprise Technology