Predictive Engineering tapped Autodesk Simulation to conduct FEA and CFD analysis to simulate air flows on a firefighter respirator mask project, as part of an effort mandated by the National Institute of Science and Technology (NIST).
Image courtesy of Predictive Engineering Inc.
FEA capabiliteis are definitely more common and more likely to be integrated with the core CAD platform. While we're starting to see more CFD integration, it definitely seems to be more specialized and tuned for specific industries while the need for FEA appears to be more universal.
Simulation software, including FEA, CFD, and solidification codes, among others, can be a tremendous resource. For example, if you are designing a cast piston for an engine, you can simulate everything from how the mold will fill and how the metal will solidify, to the stresses and strains the part will experience in service and how the dome shape will interact with the spray pattern. This allows you to refine the part for both manufacturability and performance long before it is even made.
That being said, in order to make good use of these tools, it's absolutely essential to have a solid engineering understanding of the physical situation. The increased user friendliness of software is a double edged sword. It's easy enough to set up a simulation and get numbers out - but knowing how to set the simulation up to accurately reflect the parameters of the physical situation, how to interpret and use the results, and how to tell whether the results make sense, are another matter.
One mistake I've often seen from engineers (who should know better!) is to use FEA to find the maximum stress in a part, and then compare this maximum stress to the "fatigue strength" of the material. Among other things, these engineers must have been sleeping in class when their professors were talking about the Goodman diagram. (Actually, the Goodman diagram itself is a highly inaccurate way of taking mean stress effects into account, as this paper shows).
Of course, there are fatigue codes like feSafe which apply very sophisticated approaches to fatigue. But even still, there is no substitute for physical insight into the problem.
Dave: You aptly point out one of the dangers of making simulation tools more accessible to mainstream folks. If you don't know the physics of what you're trying to simulate or set up the simulation model correctly, there is opportunity for making assumptions that lead you down a problematic design path.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.