It's because two edges can fit together like pins and tails on a dovetail and you would not detect any deviation because they fit each other, but if you add a third edge it could potentially fit the pins or the tails but obviously not both.
Printable View
Actually,you can see .0001" in a gap,not .001" as your article states,Kees. I wonder how much of a machinist the author is?
The article is just an opinion on how to do it. The great early engineering masters like Maudslay ,Whitworth,and others laid down the foundations of all modern precision in the late 18th.,early 19th.C.. They used 3 plates,scraped together to make master flats. The same applied to straight edges,though if you had already made a suitable master flat,you could use it to make the straight edge. Starting from scratch,though,you need to make 3. But,the tolerances being discussed are far beyond the requirements of woodworking,so it is moot anyway.
I'm sorry, Mike. Believe it or not, the worst offenders are engineers (or former engineers like me) and machinists. They're constantly asking you to "hand them the mics", and they're "micing" this and "micing" that....and then you ask them for calipers and they hand you dividers, which is a type of caliper but obviously not what I want! And then there's the odd duck that hands you digital calipers and decides to call them "verniers". LOL. I guess the alternate lingo was a survival skill in my former life. You're right. I'll cut it out and be more precise.
Actually,dividers have 2 straight legs with sharp tips,commonly used to scribe circles,or in many cases,machinists file one leg slightly longer. Then,they can hang the longer of the legs off of an edge,and scribe a line parallel to that edge along the metal. To be more exact,though,hermaphrodite calipers(morphy calipers) are the more correct instrument for this application. They have one pointed leg,with adjustable tip,and the other leg has a curve to bear against the side of the metal being scribed.
Being mostly underpaid,frugal types,machinists just made do with their regular dividers rather than also purchasing the hermaphrodite calipers,so you don't commonly see them. You do commonly see dividers with slightly different length legs.
Calipers are either inside or outside style. Both have curved legs,the inside being curved outwards to bear against the inside of a hole or gap being measured. I have to go eat now. More later.
Edit: So,if dividers are really compasses,why are they called dividers? Because their sharp points were used with steel rules with very thin etched or engraved divisions. A machinist was expected to be able to use his dividers to "click" into 1/128" divisions on his steel rules,then transfer the measurement to his work.
Early machinists most of the time had no micrometers of their own. They cost too much. They were expected to learn to "feel" a thousandth with plain calipers when making parts fit. This is a skill I learned early on. They sometimes had standard gauges that measured just 1 measurement,like 1",or other common measurements. They could use these to check the sizes of shafts being turned,or set their calipers to them. Much work was done by individual fitting of parts by feeling with calipers,rather than by measuring parts to absolutely known measurements with micrometers of gauges.
Here's another write up on the three edges technique: http://home.comcast.net/~jaswensen/m...ight_edge.html
It seems to me Charlie that we react at a gut or emotional level to the de-skilling/de-humanisation of work.
Which likely is an intuition we'd do well to take account of - lest we end up with the woodworking equivalent of these enormous glass and marble shopping centres (malls?) that sell little of any practical use to anybody. The cost of which pieces of bragging rights though was enormous - so we end up paying for them ad infinitum in the form of escalating prices driven by daft rents - despite a massive economic crash the prices still keep going up because the rent has to be paid off fewer customers. (i paid the equivalent of $20 (!!) for a bog standard cinema ticket in Dublin last night)
It's for reasons like this important that we stay conscious of what we are doing. It's easy to get locked into a habitual or unthinking machine/technology mindset, or into an equally blind more traditional one. In either case we stand to run into problems when the context changes...
ian
I still haven't read through that full link yet. Just posting quickly between meetings right now. I'll probably understand once I read through the details. What I don't get is if you match plane the pieces won't any deviation in flatness be identical in the two pieces and this automatically show up as being multiplied...if they're match planed I don't see how the deviations can cancel eachother out. How could they be opposing if they are planed together? I don't doubt you guys are correct. Clearly there is something I'm not visualizing correctly hear. I'll read through the whole thing on my lunch today and repost with my understanding or lack there of.
When using auger bits or handheld drills, I stopped trying to measure the dowel stock versus the auger bit - any run out in the drill being used or maybe an imperceptible bend in the auger bit might change things, or maybe I'm just measuring the auger bit wrong - I've just found it quicker to grab a piece of scrap stock from the project (often the "horns" cut off after cutting mortises, for instance) and a few auger or drill bits that might be close to what I want and drill a few holes, and try the fit of the dowels in question in those holes. Sometimes with less-forgiving wood, Irwin bit "A" might make a hole more amenable to what I want than Jennings bit "B" or even Irwin bit "C". I toss it at the end of the project, as test fitting pegs tends to wallow out the holes. (I learned that quickly after trying to rely on the same couple of pieces of drilled wood, although I do still have on kicking around as it was handy for building tool holders, seeing if a 3/4" or 13/16" or whatever is a better size hole for this chisel to rest in)
I don't know about making a straight edge via match planing. I was talking about why you need 3 pieces to make one straight edge.
What match planing buys you is if you're not perfectly square, when you join the edges the angles will always be supplementary and the glue up will be flat. It will indeed double any error in flatness along the gluing surface, however. Even worse, if there is any sanding or scraping to be done after the glueup, and the joint isn't dead square, you will end up with a wavy joint because the joint is angled and meets at different points as you go down a piece. Maybe not a big deal if you're just going to do a light scraping, but if you do any shaping, carving, rounding, etc it can end up looking really amateurish.
Just have faith that the original masters knew what they were doing. The foundations of our modern precision work is based upon them,and upon clock makers. Clockmakers were the precision industries of their day. When heavier engineering came along,the now famous early engineers like Maudslay,Whitworth,and all the others stepped in and laid the way. They scraped in 3's.
WOW,
I spent most of my life working as an Inspector which means that I already know that I don't know it all yet :)
I used to hang out in machine shops when I was in high school. Worked on every type of engine you can imagine when Uncle Sam offered me a job that I wasn't allowed to refuse so I joined the Navy. Later in life I attended a Non-Destructive Testing / Quality Control training program offered by one of Newport News Shipbuilding's (NNS) subsidiaries which led me to Nuclear Contruction projects, then Nuclear Operations and eventually ten years at NASA Langley.
I remember many visits to the Main Machine Shop at NNS and the things I witnessed few can imagine. Like OD Micrometers that were so big they had to moved by an overhead crane, used inside a BIG shop that was kept at 68 degrees F to inspect things that I am not allowed to discuss. Through the years I eventually became an ASNT Level III Examiner and racked up a lot of certifications from Industrial Radiographer to High Strength Structural Steel Inspection and last but not least Visual Inspector.
Getting back on topic I own several OD Mics, two digital calipers, a pipe wall thickness micrometer and a host of other precision tools these days simply because I use to own a metal lathe and a mill. I do have a 1.000" calibration standard that I use to check the small OD Mics and digital calipers against on occasion even though it rarely matters anymore. I never use my tape measure for woodworking, it is used for rough carpentry and to measure walls so I can calculate how much paint I need to buy.
I do build custom jigs and fixtures every now and then, thats when I am particular about accuracy. This is when the mics and calipers become very handy to own. Not that I need to work to Nuclear or Space Program tolerances per se, its just nice to be able to know that a jig or fixture won't add an unacceptable error to a project that is complex. Last but not least I own a CNC Router but it pretty much takes care of itself because I trust its accuracy based on checking the machine periodically the same as I check other tools as Stanley stated in his post.
This thread takes me on a trip down memory lane, thanks to everyone for sharing their thoughts and please remember to keep things friendly. Tolerance of other peoples personalities and the techniques they prefer are just as important as the ones you have to measure with those micrometers and those wonderful Starret scales :)
You do know that a ruler is a monarch and a scale is a measuring instrument :)
.
Stanley,not being a curmudgeon. If you are talking about technical subjects,it is an obligation to call things by their correct names. That includes calipers,dividers,micrometers,dial and digital calipers,and the accuracies which can be ascribed to these instruments. You can't throw around incorrect terms as it adds confusion to the discussion. You can't say .0001" when you should say .0005" in a technical discussion. If you work in metric,this can be confusing to convert,I know.
If scientific papers were written in very offhand ways,we'd still be cranking our cars. That's just the way it is. It is reality. Even in Japanese.:)
I didn't start this discussion,but being a toolmaker,I have used these tools for many years.
When I was in college,I was fortunate to study with one of the foremost artists in the country. I made a drawing of a sailing ship,and made some error in the rigging. He told me so,and said if you are going to make drawings of technical subjects,you can be criticized for doing it wrong. I didn't even think of calling him a curmudgeon. I knew he was right. I resolved to do better next time.
Keith,I don't use a tape for important things because acceptable government standard error is 3/16" in 6 feet! I do have a $3.00 tape measure I bought out of a bin at the grocery store. It is actually quite accurate (but not a steel rule)!!
Some mentions in this thread about some company's precision tools not being made as well now as they were before. There is a lot of good used stuff competing with new. In a relatively small market any tool not likely to get worn out is eventually back on the market and competing with current production. Capability of accuracy does not necessarily get translated into product. My son is an engineer in a company that does high precision work. They have gone to laser type measuring and recently sold off some LARGE precision tools at low prices. Purchasers got good deals.
George, please cut me some slack over a spelling error of the name of a Japanese company, the phonetic of which can be Romanized as either Mitutoyo, using the prewar Romanized version of ツ, or Mitsutoyo, using the postwar Romanized version of ツ. My mistake was to add the s, an error of spelling, but not quite as serious an error as my brother calling a vernier caliper a venereal calibre.:eek: A Freudian slip, perhaps? BTW, do you pronounce Nikon cameras naikahn, or nick-on, or nee-cone? I promise to not call you an imprecise scientist. Didn't I already apologized for my error of referring to the fourth decimal place as .0001"? Definitely curmudgeon-like.:p
Wow, I had not realized this had developed into a discussion of scientific papers and technical subjects equal in importance to the development of the modern crankless automobile! I will try to be more careful.:D
Stan
Ok, I just read through the explanation of the 3 straight edge averaging and that makes perfect sense.
RE: the match planing thing. Yes it the same thing I do when edge joining two boards but with a different purpose (though when I match plane I still do check for a slight gap when held together to create a slightly sprung joint. When I match plane in an attempt to get a straight edge the idea is that any gap I get is twice the out of flatness of each straight edge. So if my edges are slightly convex and I can't fit a piece of paper in the gap when I hold them together the gap is probably less than .002". Keep in mind I'm not using this for super high tolerance stuff. I use them as winding sticks and also to check my progress when I'm flattening a board. They seem to work well for that.
Honestly, I don't recall where I picked this up, and perhaps its a poor method but for me it seems to have been a nice quick way to get a pretty straight reference on things like winding stick that are out being used a lot and thus need semi regular truing. I like the idea of the 3 edge approach for something more precise...if I was going to do that I'd probably use a material that is more stable than wood, but easier to work than metal...I wonder if some type of hardish polymer would work well for a shop made straight edge?