22nd May 2014 by Mike Heskitt
In the optimization world, we talk of constraints to the problem. The name isn’t chosen arbitrarily, and it isn’t talking about how you support the part. It’s truly constraining the creativity of the solution you arrive at. We might think we are being clever to “drive it to the RIGHT solution”, but this is a fight the engineer needs to lose, at least in the first round of the fight.
I really like the AT&T commercial where the guy asks children around a table questions. He asked one boy if he’d rather swim in a big fun pool or a small not-so-fun pool. The expected answer is “big pools are better than small pools.” Instead, the boy asked “Does the big pool have a dinosaur that can turn into a robot and chop the water like a karate ninja?” Without missing a beat, the guy glibly replied “Of course. Why wouldn’t it?” (View the advert here).
Instead of being a ridiculous question, can we come to a place we expect our reaction to be like this for an optimization problem? “Of course. Why would we constrain it?”
OK… a lot of “buts”, lets skip those for now and just say it’s moving from a pipe dream to just a dream.
At Altair, we ALWAYS run completely unconstrained optimization as the first step of any problem. The exercise is often enlightening. Many times it surprises us with a solution we never would have conceived. Sometimes we later have to throw that idea away, but sometimes not. Always, we learn more about the problem we are trying to solve. So… regardless of how much you know about which material or what manufacturing process… before you input all of these into the problem, run it without all those constraints. See what physics wants that part to be. Imagine a way to make it.
How is this related to 3D printing? Because of the complex and unique shapes that can be produced with these machines. We need to watch this industry closely as they improve the processes and add new and better materials. I personally had mentally dismissed 3D printing because “an optimized part made from suboptimal materials isn’t optimum”. They are closing the gap quickly so I know I can’t assume this is true any more… They now have my attention.