Designing products goes well beyond simply drafting geometry for manufacturing. Aside from understanding manufacturability, it requires a thorough understanding of function, performance, and failure modes. Nowadays, Computer-aided Design (CAD) is used to draft and store geometry, as well as important manufacturing and supply chain information. While CAD is the ultimate source for product definition and documentation, it does not have any information at all regarding performance, failure, or manufacturability. We use prototypes for that, and with the increased use of computers in engineering, the prototypes have become virtual rather than physical. This use of virtual prototypes is called Computer-aided Engineering (CAE).
Historically, industry practices have limited virtual prototypes to design verification and validation, but computational simulation can serve far more of a purpose than just virtual testing. Rather than making design decisions relying on guess work, simulation can lead to informed design decisions based on both functional and performance requirements. The initial step may come off an existing design, an artistic sketch, a functional idea, or a geometric design space. Basic performance should be considered during the generative steps that precede even CAD, then, as the design progresses, the virtual prototype matures and further steers the design changes. This is truly simulation-driven design! The process is not automatic and requires the human element; there are still decisions to be made, just more informed decisions based on function, performance, physics, and aesthetics.
Design always has to relate to manufacturing and materials. New manufacturing techniques and materials have greatly extended the definition of manufacturability. For example, additive manufacturing technologies can result in significant performance gains with highly specialized designs enabled by new abilities to manufacture shapes that would be impossible to make using conventional methods. Similarly, no human can reasonably compete with a computer in the daunting challenge of designing a complex composite lay-up. By experimentation alone, simulation can rapidly process the myriad of design variants much more effectively than any physical prototyping. Furthermore, inverse computational methods can even tailor a composite material design to meet highly specific performance requirements. The proper application of simulation-driven design can master any of these challenges.
During the 1980s I was a professor at the University of Rostock in what was then East Germany. A situation arose in which the East German government shipyard had agreed to sell ships to the Soviet Union at a cost point, which proved difficult to achieve. As my work was focused on structural optimization, I was called in to see what could be done to reduce cost while maintaining performance. Using the best tools available at the time – optimization theory and hand calculations – we were in fact able to reduce the hull and stringer assembly cost. Doing so required several weeks of residence in a windowless apartment with two other professors, entering the shipyard gates each day before sunrise for another grinding bout of calculations beginning at 6 a.m. with the bleakly cold Baltic Sea air giving little hope for an expedient end to the numerical slog.
Today, given that same shipyard situation, we would quickly create a finite element model and develop a stringer structure optimized for stiffness, reduced material usage, and minimum cost. Instead of “fix it” calculations we would use simulation-driven design to achieve much better results in only a few days versus several weeks. Putting myself back in the place of my much-younger professor self, it would have been a fantastic thing to have had today’s tools available for the task at hand.
Another important aspect in both design and innovation is the dual role of design exploration. First, and somewhat self-evident, increasing the number of design variants (for example shapes, dimensions, or materials) will lead to a better understanding of the design itself which will consequently result in a better product. Second, every product maker must eventually embrace the truth of variability in manufacturing – this literally means the product coming off the line is not exactly as designed. This variability must be understood, quantified, and ultimately mitigated as part of the design itself. The amount of complexity in either of these explorations once again exceeds the capability of any battery of physical prototype testing, and must be driven with computational simulation and human reasoning.
To further innovation, simulation must achieve 3 goals:
1. The software itself must effectively support determining design direction, and drive decisions. This transfer of knowledge goes beyond simply identifying the effective design, and will also facilitate estimating the consequences of design changes throughout the process. Optimization technologies in the broadest sense are suited to provide design direction.
2. Model the complex, multi-disciplinary design physics in great detail with a correspondingly broad portfolio of modeling capabilities. One option is to capture all the phenomena into a single high-fidelity model that can range from complex non-linear, multi-physics simulations to manufacturability. However, most often performance and failure are better understood with models with fewer complexities. Therefore, the simulation software portfolio needs to consist of more than just 3D modeling, but also encompass a broader range of tools specifically targeted at performance disciplines such as signal analysis and physics-based modeling.
3. Computational efficiency requires computers and algorithms to function in unison. As computer architectures become increasingly parallel, it becomes imperative that the software running on them be scalable and robust. Simulation-driven innovation depends on massive computation to produce the data that leads to informed decision-making. User experience, computational resources and application software need to be viewed as an end-to-end solution (or appliance) to accelerate exploration and discovery; software licensing must evolve to reflect this new reality as well.
In summary, there are three enablers for simulation-driven innovation: optimization technologies; broad portfolio of physics; and computational performance. All three need to be considered and present as part of an organization’s CAE and overall product development innovation strategy. As engineers and product designers, we have a professional imperative to use the best tools possible in order to achieve results undreamt of just a few decades ago.
Find more of our stories on Simulation-driven Innovation.