Wow, it's been a whole month since I posted anything - and only three weeks until I leave, too! It's hard to believe things have gone by so quickly, especially since there's still so much to do! Okay, enough with the exclamation points, but time really has flown by.
Long Time, No See!
Brett Rowley '14
July 09, 2011
Let's start with a situation report on my research. Last time we spoke I was still working on getting the flutter jig back together. Sadly, I still don't have a new wing because the person in charge of the rapid prototyping machine has gone on vacation. As such, we got to work on the electrical systems. There are three channels to take data from: the linear displacement &angular position of the wing and the current generated in the harvester. We have a laser displacement sensor for the wing displacement and are measuring the voltage drop across a resistor for the harvester. Those are nice easy measurements akin to the measurements we'd make in Modeling & Control.
Angular position is a little harder. For that we need an optical quadrature encoder (yay for big impressive names). I had been told it was also a straightforward "plug-and-play" type sensor. A week of banging my head against a wall revealed that it wasn't. It helped me confirm that I want to be a mechanical engineer, not an electrical engineer, though. For MEs, you can spot flaws by looking at the enormous crack or the fire or the leak. Sometimes EEs can spot flaws by fire, but mostly it's a lot of poking around with a multi-meter and scratching heads. The worst part of the story is that there wasn't an impressive problem that no one would have spotted and oh-aren't-we-so-clever. No. I need a couple pull-up resistors to help get the current right. Garland (a gentleman who helps our SCOPE teams a lot) came to take a look at it for me and diagnosed the problem immediately. But in the end, all of our electrical stuff works just fine. And it's all neatly arranged and organized.
Next step was to "characterize the harvester". It's basically a spring, so we needed to know what it's stiffness (how much force it exerts per distance displaced) was and how damped it was (how long does it take it to go from maximum vibration to zero?). Sparing the technical details, I got to play with big aluminium versions of Lego and made the two jigs below.
"Custom" = "Improvised"
It turns out that business cards make great laser sights!
And then came the math. So much math. MATLAB made interpreting the data collected easy, but I'm currently in the midst of developing a model to actually put them in. There exists a nice mathematical model for flutter, but it makes some assumptions that are no longer true. Because we've added the harvester to the jig, we have to add the damping and stiffness of the harvester, too. Damping is easy, that fits into an existing part of the model. But because of the way magnets work, the stiffness of the harvester is not like the stiffness of a spring. A spring is linear - pushing a spring twice as far takes twice as much force. It turns out that our harvester's stiffness is best characterized by a 5th-degree polynomial, not a straight line. So we have to add a bunch of terms that make the math...less pretty. And then you have to re-arrange it and make the thing look like something you could solve in MatLab. As I said, I'm a Mech. E., so this is all a crazy and absurd exercise that is surely impossible great learning opportunity.
As I said, I have three weeks left before I leave, then another ten days after that before the abstract for the research conference is due. Hopefully I can get the math sorted next week, but there isn't huge pressure since we can't take any usable data until the new wing is made.
And that's where my research stands at the moment. I'll make sure to keep you in the loop as we get closer to the finish. Until next time!
Posted in: Class of 2014