More engineers than ever are relying on finite element analysis to predict, as closely as possible, how products will behave in the real world. Besides predicting and avoiding product failures, FEA can give companies the confidence required to innovate and leapfrog old design strategies. It can help reduce the number and cost of physical prototypes, allowing new products to be released faster.
FEA for the design community is also coming of age. It has entered the daily cycle of design and test and is now accessible to companies who would not have considered FEA in the past. Improved computing power, robust mathematics, and more-efficient software make this possible.
But a danger still exists in misapplying FEA tools in a design setting. Designers who are not trained in FEA must be aware that analysis software makes assumptions about the physical forces at work during product performance. Analysis software based on simplistic assumptions about the real world will not provide a reliable basis for understanding product behavior. Accurate simulation requires a sophisticated and deep approach to physical complexity.
There are no simple problems in engineering. Resist the mantra, "Red is bad, blue is good." Don't make decisions based solely on color plots. A reliable analysis workflow that accurately simulates product or process performance will not just pop out of a box of software.
Have we been here before? Remember the unquestioning trust once placed in computer-aided-design software. At one time, engineers used to believe their CAD models produced finished, perfect geometry. Only much later did designers learn about the problems that unhealed geometry and unrealistic default tolerances caused in manufacturing.
An important reason for performing simulation on a digital product model is to reduce or eliminate testing of expensive physical prototypes. Accurate simulation correlates with physical tests, increasing engineering confidence that analysis results reliably predict product operating behavior and durability in the real world. But unless the simulation accurately models real-world complexity, analysis results will never correlate with physical tests. At this point, engineering management may be tempted to throw up their hands and conclude that simulation is a waste of time.
That would be tragic. Because simulation works. One leading consumer products company recently deployed a customized system that conveniently simulates, overnight, every important force that its particular packaging designs undergo.
So how can you make sure your analysis workflow is simulating behavior that will actually occur in the life of your product? Ask some questions.
For example, how are abusive loading conditions defined? How is durability defined? Do these definitions reference the knowledge available from the team of specialist analysts and test engineers who have hands-on experience with your product?
Your analysis software should have sufficient mathematics horsepower under the hood. It should help you make the decisions that are possible for you to make and inspire a healthy respect for the complexity of everything else.
Tweet This
12 
