Information technology has transformed materials science, giving product designers the knowledge tools once controlled by specialists and turbo-charging the product-development process. When a materials research team posited in a 1989 issue of Scientific American that the Age of Materials was over -- that it would be replaced by the Era of Information -- they were only half right. Instead, the two periods collided, then fused into a new material age, one in which information technology enables manufacturers to better understand and take better advantage of material performance. Manufacturers began using simulation tools such as finite element analysis (FEA) to test how a particular material will perform in a specific application. Now manufacturers are taking advantage of the next stage of the IT and materials integration process: putting advanced simulation technology behind a user-friendly interface. With help from FEA software vendors, manufacturers are strategically and systematically bringing analysis tools into the design-engineering environment, thereby bridging two traditionally segregated domains. With this new approach, designers are able to routinely investigate alternative material/structure scenarios, which streamline design and testing and, ultimately, improve the company's competitive position. At an engineering level, obtaining and using the best available material performance knowledge is critical to product success, says General Motors Corp.'s Dave Mattis, director, Materials Appearance Center, Warren, Mich. "In automobiles -- and other sectors as well -- product success is, now more than ever, determined by what software reveals about materials and structures and how well designers are able to leverage that knowledge." Mattis says GM's math-based engineering strategy is a broad competitive effort to capitalize on IT's potential for contributing to product development. "Improving the utilization of materials has become a very important part of that IT strategy. Our target is combining the right material with the right product design and the right manufacturing process." GM's Jim Queen, vice president, North America engineering, says that over the last five years, the strategy has produced significant benefits. Simulation and analysis software became important tools for materials selection and utilization at the company. In addition to improving the materials decision-making process, the math-based approach has reduced the costs of engineering by 40% while increasing the throughput by 33%. Also during that time, the strategy has shrunk product development time to as little as 18 months, adds Mattis. By relying on virtual crash tests, GM has been able to reduce the number of crash vehicles needed by more than 85%. At a cost of $500,000 per vehicle crash test, this adds up to significant savings. To enhance the delivery of that vehicle development strategy, GM in April more than doubled its U.S. computing capacity with IBM's POWER4 supercomputing technology. "This increased computer capacity . . . will enable better decisions about vehicle design and material specifications, ultimately leading to higher quality and superior performance vehicles," says Terry Kline, global product development process information officer. Other manufacturers agree that major benefits accrue from incorporating software tools such as FEA into the product design process. The first is the capture of knowledge, workflow and methodology. The second is increased design productivity. Product design need not be a longhand process, where successful project approaches are re-created each time from memory. The use of process automation can be a sound approach to saving time and making analysis relevant to the harried designer. Design productivity is improved significantly by systematically driving FEA into the earliest phases of decision-making. Because there are no simple problems in engineering and because accurately predicting material behavior requires sophisticated analysis, auto-mating advanced FEA is now considered to be critical for many manufacturers. It allows for fast analysis, resulting in better products from better-tailored materials delivered faster to markets at less cost and higher quality. Yet to achieve these benefits, manufacturing executives must break free of conventional thinking about engineering software tools. What's important to consider is the new positioning of FEA, say industry analysts. It's no longer a comparatively pricey tool of limited functionality. FEA is emerging from being a complex tool of specialists to one that is simpler and easier to deploy throughout an organization. Typical vendors include Abaqus Inc., Pawtucket, R.I.; Ansys Inc., Canonsburg, Pa.; LMS North America, Troy, Mich.; and MSC Software Corp., Santa Ana, Calif. Winning With Packaging For leading edge adopters, the challenge is in changing internal processes to accommodate the new possibilities of evaluating materials and structures. One successful example is an automated system for testing and evaluating packaging materials and structures at Cincinnati-based Procter & Gamble Co. The goal: establish and maintain an automated FEA-based system to reduce prototyping and cut design time, says David Henning, manager of package analysis at P&G. "Every month we're tweaking old designs and evaluating new [approaches]." Package weights and sizes that are analyzed vary anywhere from a full two-gallon bottle of Tide to an empty four-ounce bottle of Old Spice. At P&G, package design proposals and revisions come from a number of sources including marketing reports, consumer surveys and ergonometric tests. Each time an aspect of a package changes, the design must run a gauntlet of analyses to prove the container will perform. In the past, P&G analysts would test physical prototypes of bottles on an oval test conveyor, nicknamed the Racetrack. The test results were reliable, but the method was time-consuming. Physical prototyping costs limited the number of designs that could be tested for any one package, thus restricting innovation. "Physical testing was accurate, but slow," Henning says. "And in our business, even 24 hours is a long time to delay a design decision." To speed things up, P&G developed a sophisticated digital system that automates the analysis of bottles for structural performance. Called the Virtual Package Simulation (VPS) system, it incorporates Hypermesh as preprocessor and Abaqus software for FEA. Because nearly all the P&G package designs share common features such as top filling, thin walls and plastic materials, the VPS system includes a number of prewritten load cases that define the load sets, boundary conditions, contact definitions and procedures for the analyses. A new feature of the VPS system is a dynamic testing system called the Virtual Racetrack. Henning describes it as a simulation of conveyor belt features, speeds and conditions. To develop it, analysts reviewed data from years of prototyping on the physical racetrack. The Virtual Racetrack gives P&G the capability of performing real-time analyses of package travel and impacts, as well as static loads. Running automated analyses on the Virtual Racetrack has produced immediate benefits. Time and cost savings are the most obvious, says Henning, but design benefits occur as well. One is the ability to run a wider range of proposed models at low cost, which promotes increased innovation because engineers can evaluate more design alternatives. Other benefits include the sense of confidence that come from the standards and uniform methods developed for the analysis, adds Henning. "We have not only reduced the time we spend modeling bottles," Henning says. "We have made our analysis process more consistent." Additionally, P&G plans to license its VPS system. At an Abaqus user conference, Henning offered this advice:
- First is simplicity. One of the purposes of automation is to make the process simple and easy to accomplish. Simplifying the process of creating an analysis is not easy. Automation is worth the time and effort for highly repeated tasks that require experts in complicated software.
- Second is speed. Stay focused on making the system fast. Minimize the number of user inputs and interactions. Justify each and every action required by the user and try to eliminate it if possible.
- Third is programming. Make sure good technical computing resources are available. These resources should be familiar not only with computer technology [such as] programming, but also have a background in mechanical engineering. They will be able to embrace the vision and see the purpose.
- Finally, maintenance. The work is not over when the automated system [for analyzing materials and structures] is delivered. Keeping the system running will require a big commitment. Automated systems area born and will ultimately die. In between they require attention and maintenance. Changes in technology will drive the need for maintenance.