ALBUQUERQUE, N.M. — Sponsored by the U.S. Department of Energy’s Defense Programs Office and cooperative research and development agreements with private industry, Sandia National Laboratories is developing a number of computer-based systems that can serve the nation’s nuclear weapons complex while fitting into the modern manufacturing world of integrated product and process design.
Sandia researchers have dubbed these technologies SmartProcess systems. These computer-based systems intelligently integrate process knowledge, product requirements, and predictive models (both empirical and numerical) within an easy-to-use interactive environment. The flexibility to modify and extend the knowledge within the systems allows for increased customization, decreased cost, and higher quality — the requirements that led manufacturers to this new engineering approach.
Integrated product and process design differs from past manufacturing design practices in that manufacturability is addressed throughout the design to manufacturing product realization cycle. Kim Mahin, manager of Sandia’s materials processing for advanced manufacturing programs, says the difference between concurrent product and process design and the traditional sequential design approach is comparable to two different types of races.
“The old way of working on the product in sequence was more like handing off the baton in a relay race,” she says. “The new way is more like a pit crew at the Indy 500, in which everyone works on the product design at the same time. Our SmartProcess program is developing the tool kits these manufacturing design pit crews need.”
Mahin is one of three Sandia leaders of the SmartProcess program. The others are Hal Morgan, who is leading the engineering sciences work, and John Mitchiner, who is leading the information systems applications work.
With SmartProcess tools, manufacturers can make rapid tradeoff decisions that balance performance with manufacturability early in the design cycle, thereby minimizing on-the-floor prototyping, which inevitably generates high scrap and rework and usually compromises function, schedule, or cost.
At Sandia, specialists in materials science, engineering, manufacturing, knowledge engineering, and computer sciences are working together to design the systems and provide the knowledge, models and integrating architectures that form the basis of these SmartProcess systems. Examples of ongoing projects include:
- FASTCAST, a system for rapid design and prototyping of gates and runners for investment casting molds.
- SmartWeld, a system for intelligent design and fabrication of welded components.
- Archimedes, a system for intelligent design and optimization of assembly procedures.
- SmartForge, a design advice system for selection and optimization of forgings for critical high pressure designs.
- SmartSolder, a model-based system for design and optimization of reliable solder connections.
- SmartProcess Advisors, which either check designs for manufacturability or provide process/material/design recommendations for manufacturing of a design.
In the non-metals area, Sandia has initiated the basic research needed to generate experimentally-based mathematical models of the physics and compaction of ceramic powders and the cure stresses generated during processing of polymeric materials.
“The SmartProcess tools integrate Sandia’s knowledge base in traditional materials and processing developed during 40-plus years of supporting the DOE weapons complex with our widely recognized expertise in process simulation,” says Mahin.
Mahin says SmartProcess applications at Sandia range from improving soldered interconnects in radars and other components to facilitating rapid designs for upgrading weapons. She says computer network links also are allowing separate DOE facilities to use SmartProcess systems in collaborative work between sites.
“One example of two sites working together with SmartProcess tools is the deployment of SmartWeld onto the manufacturing floor of Allied Signal’s Kansas City Division through a restricted network link between Sandia’s California lab complex and Kansas City,” she explains.
For industry, SmartProcess systems can lower manufacturing costs, reduce time for product realization, improve product consistency, and enhance the organizational flexibility of the design and production teams, providing value to both defense programs and domestic manufacturers.
For example, welding is a $50-billion-a-year activity in the United States, generating more than $7 billion of rework and scrap.
“The goal of the SmartWeld program is to put a sizable dent in that $7 billion figure,” says Mahin.
As an example, Mahin cites the automotive industry, which is constantly faced with trying to control distortion in welded assemblies for transmission systems. The problem is not quality control during welding, but rather quality control during forming, when residual stresses that cause the distortion develop in the material, Mahin explains.
“SmartWeld, coupled with forming process models, can predict these effects before production begins and recommend modifications to ensure optimum yield,” says Mahin. “The potential exists to reduce scrap rates and process time by factors of 100 or greater.”
Like other SmartProcess systems, SmartWeld uses solid model information (part geometry and material type) as input. With this information, users can access a knowledge-based assistant within SmartWeld to aid them in selecting the most appropriate welding process, weld joint design, and welding schedule. For new designs or materials, research-based models of process/material interactions are provided to enable engineers to rapidly identify optimum processing regimes. Interactive links to numerical process and material response models captured within three dimensional finite element simulations help designers and manufacturing engineers quantify their design margins and validate their design and schedule selections prior to committing their design to hardware.
“By providing designers and manufacturers with an integrated design, analysis, and manufacturing environment, SmartProcess enables them to concurrently optimize product performance and producibility within a desktop engineering environment,” explains Mahin. “In the future, this same manufacturing design information will be used to drive intelligent programming of automated production equipment.”
“On the manufacturing floor of tomorrow, knowledge- and model-based systems like Sandia’s SmartProcess will provide the tools for simultaneous design of products and processes. Within this virtual manufacturing environment, design modifications will be rapidly assessed in terms of cost, manufacturability, and environmental impact before any real hardware is produced,” she says.