So, yeah.. I've been away for a while.

March 17, 2011 8:13:10 PM CDT

The bulk of my 'maker time' over the last.. whew, five months.. has been spent working with metal. I made lots of chips and lots of noise, and rebuilt my lathe at least twice in the machinist's endless quest to eliminate sources of unwanted motion.

It was valuable experience, but the process isn't terribly photogenic. It's also hard to talk about because English doesn't contain the right words to build a common frame of reference between people who work with a material and people who don't.

There's nothing mystical about that.. you'd have exactly the same problem trying to describe a great pizza place to someone who's never eaten there. Words only work when they're based on shared experience, and it's hard to wrap words around a complex, nuanced, unshared experience.

So I guess you could say I've spent the last few months learning the basic vocabulary of aluminum, copper, brass, and steel.

It's an interesting language. There are lots of things you can change: cutter geometry, cutting angle, cutting speed, feed rate, chip load, type and amount of lubricant and/or coolant, not to mention the kind of material you're cutting and the kind of material making the cut. When everything lines up correctly, you get a clean, almost effortless cut. When things don't line up, you work ten times as hard for results that aren't a tenth as good.

The fun part is that there's no single right answer. If you asked ten master machinists to set up the same job, you'd get ten different solutions. They'd all look different, but they'd all work. If you came back a month later and asked them to do it again, you'd get another ten. If you quizzed one of the masters on the solutions he didn't come up with that day, you'd get responses ranging from, "yeah, that works," to, "well, I prefer doing it this way for these reasons.." and probably at least one more working solution.

The annoying part is that the solutions live in this wierd superimposed state of being simultaneously fragile and bulletproof. A small change in a single variable can turn a good solution into a bad one. On the other hand, any one of a dozen small changes to other variables will compensate for the original change, and make the solution good again.

A lot of learning to 'speak metal' involves learning how the variables interact. Some of the process is conscious, but I'd say a good two-thirds of it isn't. It's the combination of a hundred small insights gleaned from watching a good process go bad, or a bad one turn good. A huge amount of information comes in through touch, sound, and smell, and through trial and error (in my case, lots of error), you learn the signals that tell you when things are working well, and the ones that tell you something is about to go bad.

I'd say it takes five years of solid effort to get really fluent in the language of metal, then probably five or ten more to become eloquent. At five months, I've more or less stopped drooling and chewing on my toes.

Okay, I've gotten to the point where I can work comfortably to tolerances of +/-.0005". That sounds impressive if you say it fast, but really isn't all that exciting. For work that requires real precision, like lens grinding, slop like that falls in, "why did you even bother?" territory.

I've learned a lot along the way though, and that will be fodder for future articles.