When Bob the Motorized Couch needed a spine, the choice of materials was pretty straightforward. The design parameters called for five people and substantial cargo, and lumber just wasn’t going to cut it. So off we went to a metal dealer.
We pulled some pieces about the right size and shape, and then laid them across two concrete blocks. Under the dealer’s perplexed stare, we proceeded to stand on each beam and bounce up and down.
There are two kinds of engineering in this world. One kind takes place at a desk with a pencil and paper. The other takes place when you walk into Home Depot and jump on a wheelbarrow to determine if it can move the three-man rock in your yard without collapsing. A competent woodworker needs both.
Why, you ask? Why can’t you just follow the rules and be assured everything will work out?
Let me tell you another story. I was working on a team project in school. Our goal was to produce a device that could rapidly chop thick cornstalk-like grass. It had to be built in third-world countries for under $10. During our first team meeting, a teammate (call him Ziggy) made a suggestion. Ziggy’s proposal was some sort of string tied to a pole that they would spin rapidly. I was confused–we’d already determined that any sort of gearing system would blow our budget. But Ziggy didn’t intend a gearing system, in fact, he expected the shaft to be turned by a crank, effectively *reducing* the maximum rotation speed.
I was completely flummoxed. It seemed ballpeen-hammer-to-the-forehead-crease obvious to me that you couldn’t cut grass by waving string at it, but not so Ziggy. Hhe insisted on calculating the linear speed of the string, the inertial load transferred to the blade, etc. I just visualized going at a lawn Indiana-Jones style and knew this was a dead-end.
Ziggy graduated near the top of our class; he got top marks in almost all his engineering and math classes. But it would never occur to him to bounce on an iron shaft in a scrapyard as a basic design methodology.
Don’t be a ziggy. When you’re playing with a new species, run scraps through the saw, paint them with your favorite finishes, drive some screws and see if it splits, glue two pieces together and then break them apart, and generally beat on it in every way possible.
Oh, and what inspired all this: I wanted to know if pocket hole screws could hold a glue joint as tightly as clamps, so I clamped the joint just a hair less than I normally would, then drove the screw. Sure enough, a little glue squeezed out of the joint; it was definitely being clamped tighter by the screw. My little beam-bounce for the day.
To close with some words from Professor Baumgartener: “Good electrical engineers don’t blow any transistors in this lab. Bad ones blow four or five apiece. Really outstanding engineers destroy one, and they can tell you exactly how they did it.”
Go blow a transistor. You’ll thank me for it.

(The chopped grass went to feed their horses. To date they used machetes, which were fairly slow and a frequent source of missing fingers.)
(Yes, I realize that trimmers use this technique. That was Ziggy’s primary argument. But they spin at thousands of RPMs. How do I know this (he asked)? Well, I don’t know it. But both gas and electric engines runs natively at 500+ RPMs, usually higher, and gearing is expensive and tends to break, so why bother? This wasn’t good enough for him, but it turned out to be right).

(You might want to subscribe or follow me on Twitter so you don’t miss new articles)