Simulation software? Physical prototypes? Keep those licenses, but don’t clear out the shop just yet.
by Mike Rainone
 "Why in the world did we spend bazillions of dollars on these (software) programs, if you are going out to the shop to build the thing out of foam?!" |
As the focus shifts to simulation this month, it should get all of us thinking about how and why simulation has become so important to the world of engineering. Simulation’s explosive growth should be welcomed and embraced, yet you should keep that embrace at a respectful distance — like how you would embrace your mother-in-law, at least for most of us.
It seems everyone is involved in simulation, and I don’t just mean FEA. When I looked up the word “simulation,” I came to realize that my view of what constitutes a simulation had narrowed since I became heavily involved in NPD engineering. Since I’ve become immersed in engineering, I’ve come to think that simulation meant FEA, and Spice and CFD. Not so fast: from traffic flow, to the ubiquitous flight “simulation,” to blood flow simulators, to transportation simulation and modeling, to MCAD to FEA, Spice and CFD and beyond, the proliferation of cheap, powerful computing has put the mathematics and, more importantly, the understandable display of the results of the simulation, into almost everyone’s reach. (I want to emphasize that it is really the affordable display of the simulation that has driven widespread use of these programs.
ADVERTISEMENT
Think about it, just because you can run the mesh and apply the math doesn’t mean you can interpret the results - it is the visualization that makes the FEA so valuable.) Once my eyes were reopened, I began to understand what this computer revolution has spawned.
In the past, only the big R&D-heavy companies or the government could afford to implement the simulations that we take for granted today. Now almost every engineer is using simulation to one extent or another. If you have an MCAD program, you utilize simulation. If you build a 3-D CAD model, you are building a simulation — it’s not real, it’s just electrons. When you export that model into an FEA program, you have moved into a realm that was unimaginable a generation ago.
Now we have the ability to build and really test the things that we engineer before they’re built. We have the ability to test our assumptions about design before the risky dollars are spent, because each design decision is a hypothesis. It’s an experiment; it’s a theory about how that part, subassembly, device or system will work.
I don’t think we could even do our NPD work without the programming. Even our relatively small consultancy has seats of these increasingly common packages and we work on the front end of the NPD process. We don’t just use these programs after someone has done a final design; we use these tools as part of the initial, conceptual design as one might use a pencil. We use 3-D MCAD to conceptualize the parts to ask the fundamental question: “Will this concept work in all of its physical reality?”
These 3-D parts are used to feed the simulation mill, whether FEA, CFD, thermal, motion control, magnetic or multi-physics, since even at the earliest we are building math models, testing hypotheses, and determining whether our half baked ideas will work in a more efficient way than saying, “Geez, it looks like it will work.”
 An early stage COMSOL multiphysics plot of a magnetic field for a project underway at PCDworks. |
After acquiring all of this leading edge computational and visualization technology, I still find our engineers in the shop getting physical, building prototypes. My immediate reaction is to hit the roof.
“Why in the world did we spend bazillions of dollars on these programs (and if you have priced any of these packages, you know what I am talking about), if you are going out to the shop to build the thing out of foam?!” I’m sure you won’t be the least bit surprised when they say, “We really needed to see it, and play with it to understand it.”
We are now at the heart of the problem. Regardless of the veracity of the model, most systems defy true “understanding” until you get physical; until you can put it in your hands, turn it inside out, and make it work to see the interdependencies of the parts in action.
A few years ago we worked on the solution for a “lock” problem with one of our favorite customers. We held a brainstorming session on our campus, defined problems, decomposed existing designs, doodled new designs, and built parts in 3-D CAD. But because of the interdependencies of each part, we could not truthfully understand the interdependencies. So we all marched down to the prototype lab where we made parts out of steel, wood and foam. In the end, it was these physical parts that allowed us to understand how the thing really worked and helped us solve the problem.
This does not mean that I am going to turn in the licenses and give up our computer-based simulation, but I’m not going to sell off the tools and tear down the prototype shop either, at least not yet. Every 3-D MCAD package today has physics and movement built in, but until I am able to do virtual reality in which I can put on my full immersion helmet and tactile movement-enabled gloves and become TRON, we are not giving up the shop. Both have their place in the NPD process, and our job as NPD experts is to leverage both to speed time to market and reduce risk.
Mike Rainone is co-founder of PCDworks, a technology development firm specializing in breakthrough product innovation. Rainone analyzes the applications and societal implications of emerging technology on his blog—www.technologywonk.com. You can contact him via mrain1@pcdworks.com.