While I was home earlier this summer, I had to do some work on my car. Being old, it likes to break down from time to time, and being Italian, it likes to break down in the least straightforward ways possible. I have come to believe this is mostly because Italian engineers do the work of designing, but then go to the new restaurant down the street to celebrate rather than actually test anything (I mean that in the most endearing way possible). I'll spare you the technical intricacies of maintaining an old Italian car and simply say that it took three hours to remove the three bolts holding the coolant tank in place and another two hours to re-install everything (re-installation was quicker after we decided to remove the firewall between the engine bay and the rear trunk).
DFx: A Useful Thing to Consider
Brett Rowley '14
June 19, 2011
The experience taught me a very important principle: DFx. It stands for Design For X, where X is some engineering principle. Maintainability, for example, or manufacturability. It's just a way to make sure that you've met all of the relevant guidelines, not just the ones that impact you, the designer. In the case of my car, the designers certainly did not consider an owner trying to maintain the car 25 years later with household tools.
Ironically enough I've been muttering a lot about DFx while working on my research this summer. As I explained earlier, I'm not building anything from scratch, but taking two pre-existing projects and putting them together. I decided to disassemble the flutter jig to make sure everything was nice and happy and wouldn't interfere with the experiments I'll be running soon.
I wanted to remove the weights in the airfoil to weigh them and make sure they were correct. The weights are there to make sure the wing has the appropriate moment of inertia and they slot into some holes in the side of the wing. Which is great...except once in there, you can't remove them. There's no room for tools and shaking them...well, the thing is designed to shake quite a lot without trouble, so that was silly idea. Yes, I still tried.
In the end I had to break the wing open to get them out. Now, bearing DFx in mind, I wasn't simply going to make a new wing, put the weights in, and make any future researches suffer the same problem. So I drilled and tapped a hole in the side of each weight so that a bolt could be threaded into them, like a removable handle. *triumphant smile*
Anyway, Olin has a rapid prototyping machine (a couple actually, I think), so it didn't take long to get a shiny new (red!) wing made. Here it is:
What you can't see in that photo is that the new wing is the same length as the old one, but the cross section is about 10% smaller. Which is odd, seeing as I used the only appropriate file in the original CAD folder. Hmmm. This quickly became a problem for the entire jig. The team obviously made the CAD model as part of their planning stage, but made (numerous) changes as issues came up during use.
Which all leads me back to my car. I think Olin projects are like Italian cars. We come up with these plans and work tirelessly to make sure everything works for demonstration day. Sometimes everything goes according to plan, but more often than not, something comes up at the last minute. We get around it despite, over budget and out of time, but in those circumstances we're hardly concerned with whether or not someone will understand what we did a year or two down the road.
It's not a complaint - I've done that plenty of times already in my brief time here at Olin. It's merely an observation, one I'm sure that applies to schools (and companies) around the world. The moral of the story: rather than jumping in head first and just unscrewing things because that looks like the right thing to do, take the time to talk with the last person to work on the project and understand why certain things were done. You'd be surprised how many "hmm, that's weird, it didn't do that when we used it" comments you'll hear!
Posted in: Class of 2014