1. Failure to Understand the Physics
One common danger is failing to understand the underlying physics of your application. Software cannot do all the thinking for you - if you don't understand the fundamentals then disaster awaits.
The allure to play around with phenomena you don't fully comprehend is tempting. I have a vivid recollection of an engineering team that believed that they had conceived a revolutionary flow accelerator that would radically change how we capture wind energy. Unfortunately, they were using an incompressible fluid assumption in their CFD code to model supersonic flow. Let's just say that breaking the news to them was …. uncomfortable.
2. Going in Blind
Another snare is that some people expect simulation to provide predictive results of complex phenomena without experimental benchmarking or material testing. While simulation can provide valuable information about general trends, detailed experimentation is still key to achieve its true potential as a predictive tool.
Figure 1: Benefits of Simulating Early and Often
Product development teams have to make the initial investment in time and energy to carefully verify and validate the simulation method for each unique process to which it will be applied.
3. Waiting to Simulate
Perhaps the biggest mistake is to wait until the design is complete before simulating. If simulation is only performed late in the design cycle then it is almost not worth doing. Instead of accelerating the process, it actually delays the design process. You may as well take your chances with a prototype test and hope for the best. This is the main reason you hear the common cry: "I don't have time to simulate!"
On the other hand, simulating throughout the entire design process maximizes the benefits at the conceptual stage. This will drive innovation, provide guidance, allow for quick vetting of concepts, and help to avoid a late stage blind side. If performed properly, the final analysis becomes merely a confirmation. This strategy helps compress as well as define the entire process.
4. Failing to Plan
Don't forget to allocate enough time and resources for simulation when planning out the development schedule. While serving to ultimately compress the design cycle, it should come as no surprise that simulation requires manpower. Unless this allocation is prepared at the outset of the project you will likely never touch the software.
If design and simulation are performed by different groups then it is also important to collaborate closely. Teams need to adhere to timelines drawn up during the planning stages. In this way, it will be less likely that simulation will be abandoned if schedules begin to slip.
5. Assuming You Know Everything
One final trap to be aware of is failing to incorporate uncertainties as part of the simulation process. The real world is uncertain; nothing is manufactured with all dimensions being "nominal." The analysis of one geometric configuration, using one set of loads, material properties, and boundary conditions barely scratches the surface.
Figure 2: Assessing the effect of geometry changes
on component life
To perform simulation in this manner is very limiting. In the best case, you risk over-design because you need to be overly conservative. In the worst case, you experience an unacceptable rate of in-service failures.
Simulation is critical for analyzing multiple variants of a design.
With the right design process, software tools and hardware option, it is relatively easy to analyze multiple variations of the same design. Doing so will help you gain true insight into the effects of uncertainties. Failure to do this will put you at a competitive disadvantage.
Avoiding these five common traps will go a long way to ensuring your organization will perform simulations effectively and achieve all the benefits that it has to offer.
About the Author
Nicholas M. Veikos, Eng.Sc.D., is President of CAE Associates Inc. Nick has over 30 years of experience in engineering analysis across finite element theory, structural dynamics, random vibrations, structural nonlinearities, and rotordynamics.