Executive Summary

Selecting the proper materials for a structural component is critical to engineering design. Materials selection is governed by many factors, some of which are in opposition. The principal selection factors include the service requirements and design life of the product; the availability of candidate materials and the appropriate data on application-specific properties for them; the company's make or buy decision for the system components; the customer preferences; and, most importantly, the total life-cycle cost.

The use of computer-aided systems could reduce cost and design rework and requalification by providing engineering design teams with the most current materials-property data, knowledge of factors such as materials options and life-cycle costs, and available materials for a design based on experience derived from previous product developments. A Computer-Aided Materials Selection System (CAMSS) with learning capabilities would also ensure the proper archiving of materials selection decisions for future reference and accelerate the application of new materials and processing technologies by providing designers with an expanded range of possible materials and manufacturing methods for a given set of product characteristics and cost-performance criteria.

This study concentrated on the materials-specific knowledge elements of a computer-aided system. The Committee on the Application of Expert Systems to Materials Selection during Structural Design determined that the development of generic computer-aided systems is already receiving a great deal of attention within the computer science and engineering community. The basic information requirements for a computer-aided system for materials selection are receiving little attention within the materials community, however. Thus, the committee assessed that this study would have the greatest impact if it (1) detailed the capabilities required for computer-aided systems to be of value to the materials selection process during concurrent engineering; (2) identified the issues inhibiting the development of such a system; and (3) recommended materials-specific applications and developments in database, knowledge base, and materials modeling that would aid the production of a knowledge element appropriate for computer-aided systems for materials selection during design.

VISION OF A COMPUTER-AIDED MATERIALS SELECTION SYSTEM

The committee developed a conceptual architecture for a CAMSS that depicts the supporting materials-specific information technologies required. The objective of a CAMSS should be to provide design options for consideration by the design engineering team. Design and materials advisor tools should be available throughout the concurrent engineering process. Significant material properties as well as emerging considerations, such as processing and product recycling costs, will be increasingly supported by the information infrastructure. Materials knowledge should be made accessible to the engineer as reference data through design advisors that interact with product and process models that analyze, critique, improve, or optimize the design.

Major tools in the integrated environment will provide the following materials-specific capabilities: managing electronic repositories of data and documents, searching past development histories to find similar or analogous products, managing requirements, analyzing performance characteristics, modeling manufacturing and maintenance characteristics, estimating costs, suggesting improvements to the proposed product or process description, and storing the rationale for materials selection decisions for future reference. The alternative selected during concept evaluation would then be available for further refinement by the designer in a coarse-to-fine development process. To accomplish this, the CAMSS should make use of available computer-aided systems technologies. Computer-aided systems consisting of both heuristic and quantifiable design rules can be developed for subsets of the design knowledge base.

Computing technology no longer presents a barrier to the development of a CAMSS. The wide range of both hardware and software capabilities is rapidly reducing the cost of representing and implementing computer-aided



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
Executive Summary Selecting the proper materials for a structural component is critical to engineering design. Materials selection is governed by many factors, some of which are in opposition. The principal selection factors include the service requirements and design life of the product; the availability of candidate materials and the appropriate data on application-specific properties for them; the company's make or buy decision for the system components; the customer preferences; and, most importantly, the total life-cycle cost. The use of computer-aided systems could reduce cost and design rework and requalification by providing engineering design teams with the most current materials-property data, knowledge of factors such as materials options and life-cycle costs, and available materials for a design based on experience derived from previous product developments. A Computer-Aided Materials Selection System (CAMSS) with learning capabilities would also ensure the proper archiving of materials selection decisions for future reference and accelerate the application of new materials and processing technologies by providing designers with an expanded range of possible materials and manufacturing methods for a given set of product characteristics and cost-performance criteria. This study concentrated on the materials-specific knowledge elements of a computer-aided system. The Committee on the Application of Expert Systems to Materials Selection during Structural Design determined that the development of generic computer-aided systems is already receiving a great deal of attention within the computer science and engineering community. The basic information requirements for a computer-aided system for materials selection are receiving little attention within the materials community, however. Thus, the committee assessed that this study would have the greatest impact if it (1) detailed the capabilities required for computer-aided systems to be of value to the materials selection process during concurrent engineering; (2) identified the issues inhibiting the development of such a system; and (3) recommended materials-specific applications and developments in database, knowledge base, and materials modeling that would aid the production of a knowledge element appropriate for computer-aided systems for materials selection during design. VISION OF A COMPUTER-AIDED MATERIALS SELECTION SYSTEM The committee developed a conceptual architecture for a CAMSS that depicts the supporting materials-specific information technologies required. The objective of a CAMSS should be to provide design options for consideration by the design engineering team. Design and materials advisor tools should be available throughout the concurrent engineering process. Significant material properties as well as emerging considerations, such as processing and product recycling costs, will be increasingly supported by the information infrastructure. Materials knowledge should be made accessible to the engineer as reference data through design advisors that interact with product and process models that analyze, critique, improve, or optimize the design. Major tools in the integrated environment will provide the following materials-specific capabilities: managing electronic repositories of data and documents, searching past development histories to find similar or analogous products, managing requirements, analyzing performance characteristics, modeling manufacturing and maintenance characteristics, estimating costs, suggesting improvements to the proposed product or process description, and storing the rationale for materials selection decisions for future reference. The alternative selected during concept evaluation would then be available for further refinement by the designer in a coarse-to-fine development process. To accomplish this, the CAMSS should make use of available computer-aided systems technologies. Computer-aided systems consisting of both heuristic and quantifiable design rules can be developed for subsets of the design knowledge base. Computing technology no longer presents a barrier to the development of a CAMSS. The wide range of both hardware and software capabilities is rapidly reducing the cost of representing and implementing computer-aided

OCR for page 1
system logic and process simulation within affordable limits. Advances in reduced instruction set computer (RISC) chip technology allow inexpensive workstations to perform both design layout and an embedded structural analysis or materials processing simulation. Visualization techniques coupled with simulation of system behavior at many levels can be a powerful means of conveying information to design team members. The continued evolution of cost-effective, high-performance computing in conjunction with a national information superhighway infrastructure will further assist the nation's manufacturing sector in becoming more competitive in the international market-place. STRATEGIES FOR OVERCOMING BARRIERS The committee identified two main areas that are currently preventing the development of a CAMSS: (1) database and knowledge base design, implementation, instantiation, and management and (2) structural design modeling technologies. Database and Knowledge-Base Barriers The design, implementation, instantiation, and maintenance of materials properties databases and knowledge bases are integral to the development of an effective CAMSS. For example, a design engineer cannot use a system if the underlying databases contain obsolete, extraneous, unverified, or incomplete information. The committee has found that the database and knowledge base area is currently inhibited by five barriers. Standardization of databases and knowledge bases—Constructing databases and knowledge bases that contain the relevant information required for the design process and developing systems that locate and present this data are two difficult problems because of the amount of extraneous information available and the lack of standards in the content of databases and knowledge bases. To overcome these barriers, the committee recommends that (1) standards and guidelines be developed for electronic data quality, capture, storage, analysis, and exchange (following the Computer-Aided Acquisition and Logistics Support and the Standard for the Exchange of Product approaches) and knowledge base content and construction; (2) CAMSS be designed to accept a variety of database taxonomies through the use of active, "intelligent" data dictionaries that aid the identification and conversion of the contents of different databases for use in the system; (3) links between materials databases and knowledge bases be improved and computer networks for materials-specific information communication be created (e.g., an electronic Journal of Materials Selection in Structural Design, a national materials bulletin board on Internet, or a linked network of worldwide materials data systems); and (4) electronic technical assistance be provided to design teams in electronic formats. Status of knowledge capture—Methods for knowledge capture are required to enhance the lessons-learned segment of CAMSS. These include establishing knowledge-representation taxonomies, technical context standards, and techniques to update and access this information rapidly. To overcome this barrier, the committee recommends that (1) materials and computer scientists collaborate in the development of suitable knowledge-capture systems for use in CAMSS; (2) industry design teams be encouraged to establish electronic technical databases by electronic capture of all design discussions, decisions, and lessons learned in free text, spread-sheet, computer-aided design (CAD) standards, and other multimedia formats; and (3) industry design teams be encouraged to assign specific functions within the team to specialize, categorize, index, and filter the accumulated design knowledge base and locate and access other design knowledge bases. Diffuse responsibility for generating databases—The issue of how to coordinate materials developers, component users, and materials societies to generate and integrate materials property databases requires resolution. Materials suppliers predominantly leave materials qualification programs to the user because of concerns that they will be held liable for system malfunctions caused by failures and that users will only employ

OCR for page 1
materials that they themselves have qualified. Materials societies generally do not have the resources necessary for large projects. Component manufacturers typically only qualify materials for a given application and treat the data as proprietary. To overcome this barrier, the committee recommends that (1) national team efforts of users, suppliers, materials societies, and standards organizations develop integrated material qualification programs that relate to design requirements and eventual use and (2) the resultant appropriate, independently verified data be made available in a national information infrastructure to provide a realistic, initial appraisal of the advantages of a material. Disclosure of materials data—In general, companies protect as proprietary all databases and knowledge bases that contain materials properties and production-related data, such as (1) state-of-the-art information, projections, or forecasts; (2) manufacturing labor standards, rates, and price data; and (3) weight, performance, and cost tradeoff data and criteria. To overcome this barrier, the committee recommends that CAMSS be designed to assure that proprietary portions of databases and knowledge bases are fully protected. Investment to maintain databases—It is important that the information within a database be constantly monitored, verified, and updated to ensure that the best possible information is available. To overcome this barrier, organizations must (1) assign the responsibility for maintenance of databases to a centralized function, such as a data administrator with domain experts identified to act as curators of the knowledge base, and (2) provide long-term support for database maintenance once the program is established. Structural Design Modeling Technology Barriers Modeling in structural design will be an important component of any CAMSS both to provide valid details on which to base tradeoff decisions and to reduce reliance only on materials-properties databases. Modeling techniques are required for geometric reasoning, material responses on multiple scale levels, materials processing, manufacturing processing performance, product performance, and life-cycle issues such as inspectability. Modeling techniques will also be required that simulate new materials by successive extrapolation from the properties of existing materials or by calculation from first principles. The committee identified two barriers to the development of modeling. Optimization modeling—As opposed to simply showing tradeoffs between design parameters input by users, modeling techniques will be required that can suggest modifications to optimize designs and manufacturing processes. Process optimization is an important ingredient of integrated product-process design and will be used more and more in the future as the industry fully adopts concurrent engineering to reduce manufacturing costs and converge on manufacturing solutions in a shorter time. To be useful, modeling must also be done rapidly and accurately, using normal design parameters and information from multiple knowledge bases. If modeling techniques are too slow, untrustworthy, or unable to access the proper information, they will languish. To overcome these barriers, the committee recommends that (1) materials scientists and computer engineers from industry and university collaborate to develop advanced modeling techniques to reduce reliance on straight materials data, introduce expert knowledge, provide a credible basis for tradeoff decisions, and increase trust in CAMSS; and (2) materials scientists participate in basic and applied research that establishes links between materials models at several scales (e.g., atomic, molecular-crystal, cluster-grain size, polycrystal-aggregate, substructure, structure, and system). Cultural and educational barriers to implementing modeling and analysis technology—The design process is traditionally a heuristic trial-and-error approach. Increased reliance on modeling techniques requires establishing confidence that the improved design solutions can be developed in a shorter time period. Current engineering programs do not stress the importance of training in either materials synthesis and processing or computer science. For modeling and

OCR for page 1
analysis systems in a CAMSS to be useful and effective, future engineers must receive training in computer systems, modeling and analysis systems theory, and their application to the design process. To overcome the cultural and educational barriers, the committee recommends that institutions of higher learning develop interdisciplinary programs led jointly by experts in materials science and engineering, design, and computer science that (1) expose student teams to basic approaches to computer-assisted concurrent engineering design systems in order to produce knowledgeable workers with a broad understanding of the science of processing, (2) train journeymen or master technologists to use this new technology to push acceptance of process modeling techniques to the shop floor, and (3) encourage younger faculty members to collaborate with colleagues in other departments (e.g., materials science, the traditional engineering fields, and computer science) on interdisciplinary design projects and computer-assisted concurrent engineering design systems. GENERAL CONCLUSIONS AND RECOMMENDATIONS The areas inhibiting the development and implementation of CAMSS discussed above can only be overcome by a multipronged initiative with full participation and support by the integrated product development teams (IPDTs) and materials and computer scientists and engineers in the government research and development (R&D) agencies, universities, and industrial organizations. The implementation of this vision will require (1) the development of significant demonstrations of CAMSS and disseminating the results; (2) the continued expansion of electronic storage of materials information; (3) the rapid adoption and application of developing methods of computer science and technology to enhance the representation of materials design knowledge; (4) the continued development of multilevel (atomistic to macroscopic) materials processing and constitutive behavior models that reliably predict performance and manufacturability at the scale of application; and (5) the implementation of methods to address inspectability, reliability, and maintainability. Adherence to uniform computing and materials description standards in such programs is essential to the networked linking of individual tools into much larger design knowledge and support systems in the future. The committee recommends a higher level of communication among materials-specific information systems researchers and developers through a more formal electronic interchange of research information, network-linked use of computer-aided system tools, and access to electronic materials knowledge bases. Recommendations specific to developers and users of CAMSS are Government policy makers should promote (1) the development of pre-competitive R&D programs that encourage industry, university, and government laboratories to leverage expertise and knowledge to reduce the time to develop, standardize, and implement product design support systems and materials-specific information technologies and (2) the use of the information superhighway as a means for expediting the sharing of technical information and memory among federal agencies, industries, and materials societies. Government R&D organizations (Department of Defense, Advanced Research Projects Agency, National Aeronautics and Space Administration, National Institute of Standards and Technology) should promote database and knowledge base construction and standardization, design-knowledge tool demonstrations, and pilot projects as part of their future systems programs. These programs should integrate existing computer-aided system tools. Two potential ways in which this might be accomplished are to provide (1) funding for demonstration programs with creative problem solving and design concepts to teams of university faculty and students composed of computer scientists, engineering design specialists, materials scientists, and cognitive psychologists and (2) financial incentives to industry for sharing materials property data where input to public and limited access materials knowledge bases can be controlled. Industries and universities should be encouraged to collaborate in:

OCR for page 1
developing and using well-defined standards for electronic information sharing to enable selective protection of organizational private data, company proprietary data, and industry restricted data from the public domain data; improving contact between researcher, designer, and supplier on design teams; increasing rate of generation, validation, and exchange of materials data; developing powerful programs for service life prediction of structural components from materials data, constitutive models, and in-service nondestructive testing; developing models of practical significance to product development; providing materials development data in machine readable electronic format; preparing standards for knowledge representation of materials information (e.g., properties tables, graphs, and pictorial descriptions of microstructures); publicizing success stories where experienced engineers select materials showing that proper representations together with reasoning examples will promote effective material computer-aided systems development; and developing an information base on available (network accessible) materials databases and computer-aided systems.

OCR for page 1
This page in the original is blank.