October 3, 2016
Over-engineering. Warranty problems. Unmet product performance targets. Missed windows of market opportunity. A depressing list of product design process flaws that is causing more and more companies to explore increasing their use of simulation tools, and doing so earlier in the design cycle. Companies starting from scratch with simulation (I’m looking at you, people who routinely add a “safety margin”) as well as those who have long been using CAE to replace some of their physical prototypes are exploring simulation-led design.
A typical process has subject-matter expert designers and engineers working on their parts of the overall systems, simulating their components and periodically bringing everything together for a full system simulation and/or prototype. That process works, and can be very effective at speeding time to market and delivering a solid product.
But does it deliver the most innovative design possible? That’s unclear, because it depends on how much time is allocated early on for creative thinking about the design problem.
When simulation leads the design process, designers or analysts work with early stage concepts in a sandbox-like environment. They use knowns where they can and build assumptions while waiting for more details to emerge. Subject-matter experts may start with dozens of variations to see what concepts meet basic criteria; they refine these down to perhaps five to 10 options that meet more stringent requirements before settling on one or two to pursue in greater detail. Teams go from component to subsystem to system, exploring the variants that might meet or exceed the criteria. They can develop concepts quickly and relatively cheaply, and discuss them with marketing or sales, as the voice of the customer, to identify trade-offs. What emerges from this process is a CAE-based model (note, not a manufacturing-quality CAD model) that jumpstarts detailed design.
For many companies, this is a huge shift in who does what and has access to which tools. It may turn the traditional process and hierarchy on its head. After all, if analysts are usually called in at the end of a design cycle to verify fit-for-purpose, they may not be involved in the creative thinking around the design. Giving them the floor, as it were, to suggest alternatives means they need to know the design principles as well as the software technology they’re already experts in—and not everyone is up to that task. Similarly, if you now expect designers to do simulation, you’ll have experts in the product needing to become capable users of often sophisticated simulation technology—but it’s worth it.
One category of technologies that can help here are optimization tools. Depending on the vendor and type of solution, they can apply algorithms to create strong structures that use the minimum of material or optimize a design to meet a set of other targets. The criteria can be simple or complex, and they can be singular (as light as possible) or multiplicative (lightest, strongest, cheapest to source, etc.). The end result of this optimization is a frontier, a plot or table that shows the trade-offs that could be made: weighs X kilograms using material Y, and so on. Optimizations can also include cost to manufacture, cost to operate and many other factors—the criteria are up to the design team. But the more complex the optimization, the more compute power and time.
It’s important to note that no optimization solution on the market removes the human from the design process. The computer suggests alternatives that may meet the criteria but be utterly impossible for some other reason, such as manufacturability. But the computer also has no preconceived notions about a design; it doesn’t know what the last generation of the product to hit the marketplace looked like, so isn’t bound to keep that aesthetic, for example. Optimizations can come up with alternatives that human minds won’t have thought of—and that can be a very good thing.
Without giving away any secrets, I’ve seen many projects start with a mundane concept that grows into something interesting and unique because simulation helped explore how design variants would behave. The secret is doing that early enough in a design process to make a difference. Simulation-led design is an important process innovation that uses technology to boost that innate creativity.
About the Author
Monica Schnitger is the founder, president and principal analyst of Schnitger Corporation. She has developed industry forecasts, market models and market statistics for the CAD/CAM,CAE, PLM, GIS, infrastructure and architectural / engineering / construction and plant design software markets since 1999.Follow DE