It is product developer’s job to answer three questions:

• What should we make that the user will want and, more importantly, will succeed in the market?
• Can it work?
• Will it work?

What’s the difference between “Can it work?” and “Will it work?” When you build the first one, if it fits together and performs its function, you get a good feeling. This lets you know that it can work. But you might need to make tons of them. Conversely, you may only need to make a handful, but someone will get hurt if something goes wrong. There will be variation in size, material properties, friction, spring force and environmental inputs, etc. If we can convince ourselves that the design will be tolerant to these variations, then we know that it will work.

These three questions correlate loosely to three stages in product development that start after user needs and marketing requirements have been established. These are:

• Concept development
• Preliminary engineering
• Detailed engineering

In a general way, the progress through these three stages follows a hockey stick-shaped relationship between design fidelity and time.

These three steps are very unique from each other. Concept development is done by a small team considering lots of different options. Even though the team is coming up with and comparing many different things, this stage doesn’t take very long.

As we move into preliminary engineering, the pace slows and the team grows. There is a ton to do in a short period of time. Design Concepts uses a variety of simulation and math modeling tools including FEA and CFD software packages, MathCAD and SAM. Together, we call these methods “virtual prototyping.” We have also developed a number of strategic alliances with virtual prototyping experts that let us share the load and use tools and techniques that are specific to each design challenge.

As we finalize the design, the team continues to grow and the pace slows even more. Each design risk is probed and the design and documentation evolve as the virtual prototyping and physical prototyping inform the design. As manufacturers come on board and provide feedback, the refinements needed for manufacturing and assembly also feed the design engineer’s efforts.

That’s not to say that the process is inefficient. By combining physical prototyping and testing with virtual prototyping, we hope to accelerate the process by accelerating the test loops. Once we have a robust virtual model of the product, it is much quicker to probe the design for potential failures, tweak the design and determine if we have improved the situation. The schedule will expand if we have to rebuild the design in the real world each time it fails. Even worse, testing often only shows failures and not situations where you are approaching failures. These remain hidden. Virtual prototyping shows trends and “cliffs” – thresholds where the design goes from working to not working. It is even possible to estimate how many of our assemblies will fall off that cliff.

What should we make?

Let’s look at the conceptual development phase. What are we going to make? Hopefully, we know what the product has to do but we don’t know how. We come up with alternatives. As design engineers, we use CAD design and prototyping tools naturally and efficiently.

These are powerful tools that can help us see quick and tangible progress. But they can fall short if the concept is complex or if you have to make a bunch of something. That’s where virtual prototyping comes in, roughly separated into two primary forms:

• Simulation
• Math modeling

How does virtual prototyping fit into the conceptual design phase? Ideally, these tools are used sparingly to figure out the basic physics driving the concept. Once you understand the mathematical principles, you can be confident that you have chosen the right concept and not just a concept. But it’s important to make these insights quickly. The best way to do this is to keep virtual prototyping within the design team’s control. Iteration and speed are a design engineer’s most powerful tools. As we know, a small team is more nimble.

Can it work?

Once we know what we’re building, it’s time to start engineering in earnest and figure out if we can make it. At this stage, the fundamental goal is to develop the CAD model to represent all of the nominal features that are required to embody the product. Each part in the CAD model should be designed with the manufacturing process in mind. In addition to refining the CAD model, the primary milestone in this stage is the creation of an alpha prototype that gives you confidence that your design can work.

In this phase, you hope you don’t need virtual prototyping because, if you do, then something isn’t working. We rely on FEA to figure out why a part is breaking or flexing too much or a math model to figure out why something is jamming. Beyond fixing problems, though, design feeds a math model at this stage, not the opposite. Since we know we that eventually want to sell a million widgets, we know that we need to nail the details.

Now that we have created a prototype that works, we can describe the system with a nominal math model. This model will form the basis for the next phase. It is best to bring in a virtual prototyping expert so the design team can focus on adding detail to the CAD model and building alpha prototypes that will give you confidence.

Will it work?

Finally, it’s time to turn this awesome design into a commercialized product that will work. This product needs to be safe, reliable, durable and optimized for cost, size and whatever matters. There are two major goals at this phase:

• Pick suppliers and work with them to refine the design based on their preferences
• Tweak the design based on math model simulations that let you know where your edge cases are and what design elements are contributing to variation and potential failure.

By this point, your team has expanded to include both virtual prototyping experts and your suppliers. With a larger team comes increased communication and collaboration and usually slower progress. But any time that is taken to simulate variation is much quicker than making changes once a product is being made in high quantities or optimized by prototyping and testing over and over.

In the end, all of your foresight pays off as your virtual prototyping expert performs Monte Carlo simulations and describes accurately the way variation can lead to failure and alternatively how design tweaks can help avoid these failures. By combining design, simulation, math modeling and physical modeling at the right time and with the right team, you can now confidently deliver your product!

Now it’s time to come up with your NEXT big thing.