Flutterby™! : a flaw of metric 2007-02-12 05:07:19.71716+00

a flaw of metric

2007-02-12 05:07:19.71716+00 by Dan Lyke 4 comments

One of my recent acquisitions for woodworking is a good ruler marked out in 64ths. My tools are largely metric, so what I'd really like is a ruler marked out in tiny metric increments, but here we run into a problem.

A millimeter isn't accurate enough to do much of anything. Well, okay, you can do carpentry with millimeters, but not woodworking. So here's the problem: A tenth of a millimeter is at the limits of human perception. What you're really looking for is maybe a third of a millimeter, perhaps even a quarter of a millimeter, but as soon as you start subdividing by things other than 10 you start to show up the silliness of the metric system.

And that silliness is that the metric system uses scales that aren't useful to human physiology. A quarter of the circumference of the earth (and Eric can tell you how dubious a measurement that is) divided by some power of ten is arbitrarily stupid, but using base 10 as the basis for all other measurements is immediately dated. It makes no sense in computer terms, and it makes sense in human terms only because we've gotten used to base 10 arithmetic, but not because dividing dimensions (or anything but currency) by 10 makes any sense at all.

I suppose I could just get a vernier caliper that measures tenths of a millimeter, but rulers are handier than calipers. So I'll continue to get close, and then do conversions from my measured units to metric and adjust the tools that way for the fine tuning.

Tell me again why we're abandoning something God gave the British for a system invented by a Frenchman?

I've always found that this Java animation puts "orders of magnitude" and "powers of 10" into a visually understandable perspective:

Having said that, there are of course many instances where powers of 10 have little practical meaning.

#Comment Re: Digital calipers made: 2007-02-12 20:56:05.48376+00 by: m

Some digital calipers and analog dial indicators are rather inexpensive these days. The cheaper ones are often fine for woodworking because they usually offer good precision (reproducibility) if not great accuracy. Some are available in fractions, but I am not familiar with those. Digital calipers in the $20-50 offer measurement and conversion in both metric and inches. Useful quality dial indicators can be had for $10 and up.

It can sound silly to use digital calipers and dial indicators in the wood shop, where wood movement can easily overwhelm the accuracy of any such device (and rulers as well), but they certainly have a place. I have previous experience with such instruments from some low level machine shop work when I was a student. I had a metric vernier caliper around because I also do some lapidary work where tolerences as low as 0.05mm were required for the gross measurements. I started finding applications working with wood, but the metric conversions were a pain. Some years ago I picked up a cheap 6" plastic digital caliper for about $40 (cheaper now) that I use for both wood working and gem cutting.

As I age, my eyes do not always see the markings on steel scales as easily as they once did. Especially for those graduated in 64ths or 100ths, and when the scale is in glare or just physically awkward to read. I also end up with fewer transcription errors with the digital readouts.

Tuning, setup and checking of equipment (primarily stationary) are obvious applications. It is nice to be able to check a tenon that I cut for a mortise using the inside and outside measurements. Differential measurements are great if there is a zero reset. Saves the arithmetic.

Certainly not a required item. I suspect really good woodworkers would have little need of such except for their tools, but I find a couple of precision instruments worth the small cost.

I've got a decent metal vernier caliper (because I'm old skool) in inches, but, yeah, I should probably just go spend the $40 it'd take to get a basic digital caliper in mm. I've resisted so far because it seems like for most of the stuff I'm measuring, calipers are the wrong form-factor: For router bit depth, position of the saw guide rail from the edge of a piece, distance of the guide rail or template from a joint edge, calipers don't do that well: Even with the inside fingers, they're not all that good for measuring a flat edge (the rail position or tool base) to a negative edge (the edge of the board, or the end of the cutter).

On accuracy... One of the things I'm having to get over as I move up from 2x4s and cheap birch ply cut with a Skilsaw is that if the grain is parallel and the joint is finished well, the motion of the wood isn't going to be working against the joint, and even in cheap pine people can see a 64th on a dovetail joint. Yeah, design for expansion and contraction, and there will always be a need for floating pieces, but there's nothing at all wrong with getting joints that won't be working against each other tight to the limits of the wood grain.

On contraction of the wood: We've been playing with some madrone that we cut down a year or so ago, and on a whim polished up some slabs of that and tossed 'em in the food dehydrator. The spiderweb of internal cracks that showed up in those slabs was fascinating. I'm thinking of dying some wood putty or a soft resin black, drying an end-grain slab, filling the cracks, and then polishing it so those spidery lines show up in the reddish wood. Not sure I'm up to the inlay of those into something, but I think they'd make a beautiful inlay.

Duh. Broke out the caliper last night and realized it has a poker-out-the-end for depth measurement, and has metric and imperial scales. Not great for router depth adjustment on the router itself, but if I cut a dado and measure that depth it's okay.