Precision Tools in Woodworking

[Dave Cav]

I definitely use precision tools for setting up machines. I have a Delta thickness sander and used a dial caliper to get the table exactly parallel to the sanding drum. I ran 8" wide boards through,and checked their opposite edges for thickness,adjusting the table until I got the boards exactly the same thickness all the way across. Then,when I glue up guitar tops and backs,they are exactly the same thickness,rather than looking like tapered siding.

I know this is the Neander forum, but the thread has been hijacked a little, so I'll keep the hijack going. I'm far from a 100 % Neander, as a glance at my avatar will tell you, but I do a lot of final fitting, trimming and finishing with hand tools. I use my power tools for heavy, repetative work, or some types of work that require machine-like precision- such as making mortises and tenons. And in that respect I'm kind of a machine Neanderthal. I use a flat belt Millbury tenoner that's considerably older than I am, and a Newman mortiser that was made when I was in elementary school. After I use a marking gauge to lay out the locations of the M&Ts, I use a digital caliper and a couple of dial indicators to set the machines up. After measuring the position and offset of a test mortise or tenon with the calipers, I use a dial inidicator to move the mortiser table or tenoner heads the exact amount to center the mortise or tenon in the stock, and I can usually get everything centered and ready to go in one iteration. Once the frame, or door, or whatever is made and glued up it's usually back to the chisels and hand planes for final fitting.

[ian maybury]

I'm another engineer that has posted before on the usefulness of precision in woodworking.

We've just had a thread which discussed the ability of a finely set plane to remove a shaving as fine as half a thou, and when we creep up on fits we're often dealing with dimensions if not this tight then still pretty damn fine.

Wood is elastic, and wood does move - but there are lots of places in joints where we need to generate fits to what are pretty much engineering tolerances. The wood of course moves afterwards, but the whole point is surely that the piece is designed so that these joints move with it and don't open up.

Working to fine tolerances in engineering is not a simple matter either - even on highly accurate machines. In the case of engineering parts and assemblies end up being designed around the capability of the machines that make them, and the tolerances required for them to function. Get it wrong and it'll end up a mess. Try for example marking and drilling a line of 20 equally spaced holes measuring from one to the next by whatever means and see how many ways there are that errors can cancel and/or accumulate in a situation like this. (different results that this can deliver) The issue may not arise if you are using a fancy CNC controlled machine that avoids the problem by having a positioning system designed so that it continuously positions from a datum - but that doesn't alter the underlying reality...

There are to my mind two basic methods possible in woodwork. Either (a) measure approximately, recognise this by making many parts slightly oversize, and when cutting these pieces and sneak up on fits for everything that needs a close fit - or (b) locate/size everything by measurement and proceed by dead reckoning/calculation and cut to size. My guess is that in woodworking most of us (a) evolve ways of working which actually blend the two e.g. square some critical corners and set some key lengths, but maybe fit most of the rest or reference off other parts, but that (b) these routines (because some will work better than others, and because stuff like machine accuracies can be highly variable) are not often formalised or well communicated - we work them out for ourselves.

There's by definition scope for a fairly wide range of approaches - even if we're shooting for similar outcomes.

There tend on the other hand in engineering to be fairly highly developed methodologies for the management of tolerances, dimensions and fits - and of course machine tools are often (a) capable of high levels of accuracy, but also (b) tend to be pretty clearly characterised in terms of their capability. (using the quality control and instrument calibration procedures already described)

Wonder if anything similar is taught on say the high end fine woodworking courses taught in the prestigious schools? It matters, because there's by definition going to be more and less effective and efficient methodologies. Skill/what you have practiced is of course a very big factor too - both in machine and hand tool use. There's stuff that can be done very efficiently with hand tools provided you have mastered the hand skills (put in the 1000s of hours needed to properly embed it in the body memory), and many woodworking machines can become highly inaccurate if handled wrongly.

Back in the day (pre mass production - the days of craft production) engineering was much like wood working. Parts were hand fitted, and seemingly the same parts in machines etc were not for this reason interchangeable. Economics and competitiveness drove the situation in manufacturing and engineering - we want to be able to repair products, but it also shook out that when volume production is involved that it's cheaper and delivers much higher quality levels in the finished item if you work to controlled tolerances and known and often very high levels of dimensional predictability/repeatability. A big part of the problem in DIY/hobby woodworking is no doubt that except in the case of very high volume operations the value simply isn't present to deliver a return on the required investment in improved methods and capability)

It'd be nice in many ways to see these issues discussed in factual terms as they relate to woodworking - the reality of what's 'best' (what does best mean?) is surely a lot more nuanced than might seem to be the case??

One issue that gets up my nose is seeing the makers of quite a lot DIY/hobby level woodworking tools and machinery exploit this lack of clarity/lack of understanding of how tolerances translate into performance by setting enormously wide and self serving tolerances/specs for stuff like machine table flatness for sample. There's rarely too many hostages to fortune created by these companies, but heaven help anybody that gets a lemon/a machine at the worst case/outer limits of the specified tolerances in multiple cases.....

ian

[Charlie Stanford]

I'm another engineer that has posted before on the usefulness of precision in woodworking.

We've just had a thread which discussed the ability of a finely set plane to remove a shaving as fine as half a thou, and when we creep up on fits we're often dealing with dimensions if not this tight then still pretty damn fine.

Wood is elastic, and wood does move - but there are lots of places in joints where we need to generate fits to what are pretty much engineering tolerances. The wood of course moves afterwards, but the whole point is surely that the piece is designed so that these joints move with it and don't open up.

Working to fine tolerances in engineering is not a simple matter either - even on highly accurate machines. In the case of engineering parts and assemblies end up being designed around the capability of the machines that make them, and the tolerances required for them to function. Get it wrong and it'll end up a mess. Try for example marking and drilling a line of 20 equally spaced holes measuring from one to the next by whatever means and see how many ways there are (different results that this can deliver) that errors can cancel and/or accumulate in a situation like this. The issue may not arise if you are using a fancy CNC controlled machine that avoids the problem by having a positioning system designed so that it continuously positions from a datum - but that doesn't alter the underlying reality...

There are to my mind two basic methods possible in woodwork. Either (a) measure approximately, recognise this by making many parts slightly oversize, and when cutting these pieces and sneak up on fits for everything that needs a close fit - or (b) locate/size everything by measurement and proceed by dead reckoning. My guess is that in woodworking most of us (a) evolve ways of working which actually blend the two e.g. square some critical corners and set some key lengths, but maybe fit most of the rest or reference off other parts, but that (b) these routines (because some will work better than others) are not often formalised or well communicated - we work them out for ourselves.

There's by definition scope for a fairly wide range of approaches - even if we're shooting for similar outcomes.

There tend on the other hand in engineering to be fairly highly developed methodologies for the management of tolerances, dimensions and fits - and of course machine tools are often (a) capable of high levels of accuracy, but also (b) tend to be pretty clearly characterised in terms of their capability. (using the quality control and instrument calibration procedures already described)

Wonder if anything similar is taught on say the high end fine woodworking courses taught in the prestigious schools? It matters, because there's by definition going to be more and less effective and efficient methodologies. Skill/what you have practiced is of course a very big factor too. There's stuff that can be done very efficiently with hand tools provided you have mastered the hand skills. (put in the 1000s of hours needed to properly embed it in the body memory)

Back in the day (pre mass production) engineering was much like wood working. Parts were hand fitted, and seemingly the same parts in machines etc were not for this reason interchangeable. Economics drove the situation in manufacturing and engineering - we want to be able to repair products, but it also shook out that when volume production is involved that it's cheaper and delivers much higher quality levels in the finished item if you work to controlled tolerances and known and often very high levels of dimensional predictability/repeatability.

It'd be nice in many ways to see these issues discussed in factual terms - the reality of what's 'best' (what does best mean?) is surely a lot more nuanced than might seem to be the case?? One issue that really gets up my nose is seeing the makers of quite a lot woodworking tools and machinery exploit this lack of clarity/lack of understanding of how tolerances translate into performance by setting enormously wide and self serving tolerances/specs for stuff like machine table flatness for sample. There's rarely too many hostages to fortune created, but heaven help anybody that gets a machine at the worst case/outer limits of the specified tolerances in multiple cases.....

ian

Well, here's a set of facts: The finest furniture-making, joinery, instrument making, etc. were all accomplished by hand in an age where none of these high precision measuring devices, machinery, etc. were available. Walking into just one, literally just one, Medieval cathedral should disabuse anybody with a notion of necessity for any of these things. And then after visiting just one cathedral stroll through a major museum with a decorative arts collection if for some strange reason one isn't convinced.

[ian maybury]

Hi Charlie. I guess it's not about either/or - or to argue that precision wasn't possible back in the day. Fine furniture from centuries ago is another example of what you describe. The modern tendency to de-skill stuff by using high precision but horribly expensive machines tends to produce horribly lifeless stuff, and is anyway not economically feasible unless the volumes are pretty high.

The trouble is though that even at the DIY/hobby level, and even working with hand tools (never mind in mixed hand/machine work) we do have to take account of issues of precision and dimensions, and the issue seems not to be clearly treated in a lot of what is published...

ian

[Mike Siemsen]

I think people should work the way they enjoy working. It is important to remember that planes make things straight and that you can plane out a straight edge for free and make you own squares from wood. Straight edges and squares can be checked against themselves. in hand work it is far more important to always lay out from a registration face and edge than it is to have ultra precision tools. The object is to minimize inaccuaracies through technique. I can build a perfectly square frame with out a square. equal length sides, equal length ends, measure the diagonals with a stick until they match. It is better to have a goodknowledge of geometry than precision tools.

[Jim Palmer]

Well, here's a set of facts: The finest furniture-making, joinery, instrument making, etc. were all accomplished by hand in an age where none of these high precision measuring devices, machinery, etc. were available. Walking into just one, literally just one, Medieval cathedral should disabuse anybody with a notion of necessity for any of these things. And then after visiting just one cathedral stroll through a major museum with a decorative arts collection if for some strange reason one isn't convinced.

Very true.

-----------

The need for accuracy tends to be as variable as the grains of sand on a beach until filtered down and graded depending upon end-use. Aesthetic beauty, finesse and micron-fine accuracy don't necessarily run hand in hand and one rule of thumb a carpenter or mason needs to bear in mind is the need for pieces to be pleasing to the eye. Think of the golden ratio and Michaelangelo's "David" combined. If a finish piece looks and feels right, it is right. Not necessarily the case in terms of engineering, due to the fact most pieces are intended for much higher stress applications involving extremes seldom suffered by timber goods.

Engineering and woodworking tolerances tend to diverge greatly once we consider the actual need for practical working methods and tooling tailored toward crafting such different materials.

Yes, accurate and well made tools are a must within both industries, but the need for accuracy depends greatly upon the skills, tooling, materials and end-uses involved. Are such high demands for accuracy absolutely necessary for those involved in amateur woodworking? No.

Final tool choice/selection is certainly up to the craftsman, but materials and end-use dictate the degree of accuracy needed. In terms of woodworking, micron-fine accuracy is quite unnecessary unless crafting wooden clocks in the style of Harrison's fine wooden timepieces.

By all means invest in whatever tooling is financially practicable and beg, borrow and steal techniques from other crafts for use in woodworking, but no amount of financial outlay will ever replace the very genuine need for one to learn the necessary differences in tooling and materials when crafting wood or metal. Then learn which is more important...... Actually crafting wood and enjoying the experience, or determining whether or not you need to tool up and attempt to craft to much finer tolerances than ever necessary.

------------

[Mel Fulks]

Accuracy is always relative .It seems to me that whenever something more accurate is introduced someone else involved in the manufacturing process 'uses up 'that advance by slacking off on their part. Enthusiasm about the high tech new stuff can cause people to pay more for less. Example: more accurate machining in ,Mechanite out. Lower quality new miracle product. Accuracy can can cause some to look at thousandths and ignore larger increments.I once saw an employer ecstatic over the performance of his new CNC router. A big pile of plywood parts had been cut out EXACTLY one inch too small. Unfortunately computer genius had programmed router cuts on wrong side of lines. But the parts were ACCURATE ,though unusable.

[Bob Warfield]

This thread has been a very good read for me. Here's my story. I used to build drag racing engines. When you take an engine design originally made to run 4,000 to 5,000 R.P.M.'s and build it to turn 10,000+ R.P.M.'s you have to have ALL the deminsions not close but PERFECT. A very old friend who retired from Sunnen Mfg. once proved to me that touching a part with your fingers will change it's deminsion. I got burnt out on the raceing game several years ago and haven't touched an engine since.
When I got into wood working I went nuts trying to deminsion lumber. I've since learned, almost, that real close is good enough.

Good Luck
Bob

[Ed Looney]

I have shared my opinions regarding the usefulness of precision tools for woodworking on a couple of posts. Some dismissed the usefulness of precision measuring tools without providing reasoned explanations. But I think others might benefit, so a few of my experiences and thoughts follow.

I suggest that woodworkers that desire good results and value their time should own three tools: 1. A diemaker�s square; 2. A precision ground straightedge; 3. A precision micrometer (1� .0001�). I will explain why below, but first you need to understand what kind of tools these should be, since there is so much useless junk made in Taiwan, China and India, and even the US and Europe. The tools I am recommending are not sold at Home Depot, or even in the woodworking tool catalogues, but by retailers that specialize in supplying high quality machinist�s tools to professionals. These tools are expensive, and they come with real certificates of testing for accuracy, and a solid guarantee. Once you have your set of these tools, check out their equivalents made in China, Taiwan or India and you will immediately see what I mean about junk.

There's a lot of misinformation on this thread. I routinely work to a couple thousandths of an inch without doing any special at all.

Anytime this topic comes up, there's always this broad brush stroke of, "Oh, you're just being silly...tee hee hee....the wood's going to move. What a waste of time". It's a bit more nuanced than that. Yes, at some level the wood does move, and at another level it just doesn't. In some places it pays (or is necessary) to be extremely precise, and in other places it doesn't.

Stanley I think you are spot on and John your observations on this thread are insightful.
The first error some are making about the use of precision tools is that they have to be used in daily production. I do not believe that is what was intended. There is a need to have a set of measuring tools that are not used to make anything but are only used to set-up and inspect. I have a good try try-square I love to use while laying out work but I also have a machinist grade square that I use to inspect my other squares. Rob Cosman is fond of saying that you must inspect what you expect. He was taught this by someone who taught him. But no matter the saying's origin there is nothing more true in life.

I am a machinist by trade and I was taught as a young man starting out that if I need to be accurate to 1/16th then I need to learn to cut to the 1/32nd. If the boss expects you to cut to the 1/16th and you can cut to the 1/32nd you will never have a quality issue. That was some of the best advise I was ever given.

There is a fallacy of thought that precision with wood is not possible because wood moves with temperature and humidity unlike metal. The inference there is that metal does not move which is inaccurate. Metal does move. So why the accuracy in machining metal? The answer is fit and quality of product. Metal cut at 70 deg will return to that size at 70 deg. Like wood will return to the size it was cut when the moisture content and temperature return to the same as when the wood was shaped.

I ran a test with a tenon cut from Poplar. It was cut it exactly 1/4" (.250"). I left it in my shop that is not temperature controlled and measured it throughout the next year. The maximum movement I got was .003. That is less than the average human hair. I can take 1/4" aluminum bar stock and lay it out in direct sunlight and get more movement than that.

I am a firm believer that the accuracy in woodworking can be taken to the next level if only people will choose to master the skills necessary to get there. The only obstruction is desire.


Ed

[Stanley Covington]

I have found many of the comments well thought out. Some not so much.

It was suggested that I provide links and names to manufacturers of the machinist’s tools I recommended. My straightedge and micrometer are by Starrett and made in the USA. The diemaker’s square is SPI made in Switzerland. I have been doing this for a long time, and I bought these tools well before anyone even thought about having true precision tools made in China or India. Since then I have read bad things about the quality of SPI, B&S and other stalwart name brands. I even doubt Starrett nowadays, so I will not recommend any one brand. The key thing is to buy from a reputable retailer of professional grade tools, someone with a solid money-back guarantee. Have them check it in the store in front of you against the sort of expensive certified calibration tools a high-quality machine shop would typically buy for QC purposes. You can’t do that online or from a catalogue. If they refuse to check it for you, or won’t accept returns, go elsewhere.

Allow me to clarify a few points my initial post left muddled. First, I am not suggesting your work is crappy unless you use high-precision machinists tools for your everyday woodworking. As I have mentioned in other posts, I seldom precisely measure anything beyond overall exterior dimensions and thickness.

An example of my workday tools. I am currently using an Empire aluminum square I bought on Guam for big stuff, a small cheap stainless Shinwa square I bought in Tokyo recently for small stuff, and 1.5 meter and 50 cm Shinwa stainless straightedges certified compliant with Japan Industrial Standard Grade 1 (Shinwa makes a damn good straightedge, BTW, but I have never seen them sold outside of Japan). I also have a fairly new Starrett combo square I use infrequently. These are my everyday working tools, and I imagine they are very similar to your own. I use these relatively cheap tools for two reasons. First, I know the squares are disposable, will surely be shaved by marking knife to uselessness, or will eventually get knocked off the bench or otherwise damaged, so there is no point in spending a lot of money on them. Second, I checked them for precision with my “Standard” machinists tools when I first bought them, and I also periodically check them against my “Standards” to ensure they are still nuts on, so I use them with confidence. When they don’t pass muster anymore, I will toss them or give them away. Not a matter of if, only when.

But I seldom use the division marks on these tools. I use them mostly to layout and/or confirm straight and square. Besides these tools and carefully marked layout sticks, I rely on two Starrett compasses, good marking gauges (Titemark and Kinshiro), a sharp marking knife (Kiyotada), and Mark-1 eyeball for most work. Sadly, there was no certification or guarantee on the Mark-1, and I find it is not what it once was.

I badmouthed my LV straightedge because the one I bought some years back was made in India and abysmal. They don’t appear to have it in their catalogue anymore, but if you drop by the house, you can check it out yourself.

A word about quality perception. I have a very Japanese mindset when it comes to quality: When I pay money for a product, I believe I OWN the reputation of the manufacturer and retailer, for good or ill, as far as that product goes. If the product turns out to be bad, I will tell everyone I know about it because I have paid for the right to do so. In Japan, the buyer is God. Vengeance is mine. This is not the American mindset, and many of you will object and call this unfair. Ergo Detroit getting its ass kicked over and over and over again. LV screwed up when they let a poor quality product travel from their shipping office to my front door, but it appears they are trying to do better.

So you think that wood is too unstable or too yielding for precision work? I heartily disagree. While not as homogenous as steel or concrete or plastic, wood is an excellent material that can be worked with great precision. Just ask anyone that works it commercially under a decent QC program. Wood does not deform as much as rubber, nor does it compress like a marshmallow. Do you really imagine in your feverish dreams that, because wood compresses under stress, it cannot be cut/milled to high tolerances? Wow, imagine that, a material that exhibits plastic deformation and a hysteresis loop; how unique (heavy sarcasm intended).



If you truly believe that seeking precision is futile because wood is softer than steel, you need to pay closer attention to the world around you. Extremely high tolerances in wood are not just possible but are absolutely required in many situations. One example, and not an unusual one. In a previous reincarnation I worked for five years for a company named MetalFit (http://www.metalfit.co.jp/) making wooden structural frames from softwood glulams using fully automated CNC production lines in Japan and Minnesota. The company has changed hands several times since then, but judging from the website, the connectors and production lines appear to be the same. These robotic lines cut/milled the members using programs automatically generated from CAD structural drawings. Because the wood had to accommodate precision metal connectors and drift pins to transmit compression, tension, bending and shear, cutting and milling tolerances were +- .0008". The members were all pre-cut in the factory and assembled on the jobsite, so dimensions had to be very tight to prevent Murphy's Law of Accumulated Errors from making assembly difficult. Of course, the glulams were re-milled upon delivery to bring them up to acceptable tolerances before they went into the line.

Cutting wood to such tolerances is easy if the machinery is well designed and maintained and setup properly. Anybody in the industry knows that. For that matter, so long as its arbor is running true and the blade doesn't wobble too much, your tablesaw can do it. But believe me, only insanity can result from attempting to setup your tablesaw to cut those sort of tolerances if your square is .012” out of wack like the Starrett combo I rejected was.

The Germans do a pretty good job at precision machining of wood too. Fascinating machines. http://www.hundeggerusa.com/

But let's set half-baked engineering theory, and willful ignorance of modern commercial woodworking practice aside for now, and shift our focus back to handwork. Do you suppose that a tenon joint layed out with a sloppy square, perhaps resulting in a gap at the shoulder, is the best anyone should strive for? If you are happy with that sort of joint, blissful art thou. I prefer to use a square I know to be accurate (no, not a diemaker's square, dagnabit, pay attention), and if the shoulder turns out to be sloppy, well then that's may fault for not cutting to the line. I like Sean’s words about him disappointing his tools. No doubt we all do that to one degree or another. But just because I can’t cut by hand to the ten-thousandth of an inch my micrometer can measure, must I be satisfied with +- .012” levels of accuracy? Of course not. The golden mean lies somewhere between these two extremes, and each of us makes a decision, intentionally or no, about what that precision will be. My intention was to encourage you to ask yourself “how do I know my tools are accurate?” From that starting point, you can make an intelligent decision.

I am fine with folks that don't need or don't want their layout and measuring tools to be straight and square. I could not care less if their little projects are lop-sided. But (here’s the part where I puff up and turn red) to spout half-understood materials science principles to support the notion that precision in woodworking is futile is not to talk about the "real world," but rather about "Wayne's World."

[Matt Lau]

Precision is precision.

While wood is an unstable medium, precision matters!
I had an eye opener last week when I was prepping mahogany for a neck.
My maestro wasn't satisfied until both surfaces were dead flat--they literally sucked onto a precision milled, granite plate with just rubbing!

It doesn't matter that things can be fixed/tweaked!
Sometimes it's nice to do things right the first time.

-Matt

ps. Can you please, please, please send me links to the manufacturer/suppliers?

[Stanley Covington]

I have found many of the comments well thought out. Some not so much.

Matt asked me to provide links to manufacturers of the machinist’s tools I recommended. My straightedge and micrometer are by Starrett and made in the USA. The diemaker’s square is SPI made in Switzerland. I have been doing this for a long time, and I bought these tools well before anyone even thought about having true precision tools made in China or India. Since then I have read bad things about the quality of SPI, B&S and other stalwart name brands. I even doubt Starrett nowadays, so I will not recommend any one brand. Fowler might still be OK. The key thing is to buy from a reputable retailer of professional grade tools, someone with a solid money-back guarantee. Have the store check the tool you are considering buying in front of you against the sort of expensive certified calibration tools a high-quality machine shop would typically buy for QC purposes. You can’t do that online or from a catalogue. If they refuse to check it for you, or won’t accept returns, go elsewhere.

Allow me to clarify a few points my initial post left muddled. First, I am not suggesting your work is crappy unless you use high-precision machinists tools for your everyday woodworking. As I have mentioned in other posts, I seldom precisely measure anything beyond overall exterior dimensions and thickness.

An example of my workday tools. I am currently using an Empire aluminum square I bought on Guam for big stuff, a small cheap stainless Shinwa square I bought in Tokyo recently for small stuff, and 1.5 meter and 50 cm Shinwa stainless straightedges certified compliant with Japan Industrial Standard Grade 1 (Shinwa makes a damn good straightedge, BTW, but I have never seen them sold outside of Japan). I also have a fairly new Starrett combo square I use infrequently. These are my everyday working tools, and I imagine they are very similar to your own. I use these relatively cheap tools for two reasons. First, I know the squares and straightedges are disposable, will surely be shaved by marking knife to uselessness, or will eventually get knocked off the bench or otherwise damaged, so there is no point in spending a lot of money on them. Second, I checked them for precision with my “Standard” machinists tools when I first bought them, and I also periodically check them against my “Standards” to ensure they are still nuts on, so I use them with confidence. When they don’t pass muster anymore, I will toss them or give them away. Not a matter of if, only when.

But I seldom use the division marks on these tools. I use them mostly to layout and/or confirm straight and square. Besides these tools and carefully marked layout sticks, I rely on two Starrett compasses, good marking gauges (Titemark and Kinshiro), a sharp marking knife (Kiyotada), and Mark-1 eyeball for most work. Sadly, there was no certification or guarantee on the Mark-1, and I find it is not what it once was.

I badmouthed my LV straightedge because the one I bought some years back was made in India and abysmal. They don’t appear to have it in their catalogue anymore, but if you drop by the house, you can check it out yourself.

A word about quality perception. I have a very Japanese mindset when it comes to quality: When I pay money for a product, I believe I OWN the reputation of the manufacturer and retailer, for good or ill, as far as that product goes. If the product turns out to be bad, I will tell everyone I know about it because I have paid for the right to do so. In Japan, the buyer is God. Vengeance is mine. This is not the American mindset, and many of you will object and call this unfair. Ergo Detroit getting its ass kicked over and over and over again. LV screwed up when they let a poor quality product travel from their shipping office to my front door, but it appears they are trying to do better.

So you think that wood is too unstable or too yielding for precision work? I heartily disagree. While not as homogenous as steel or concrete or plastic, wood is an excellent material that can be worked with great precision. Just ask anyone that works it commercially under a decent QC program. Wood does not deform as much as rubber, nor does it compress like a marshmallow. Do you really imagine in your feverish dreams that, because wood compresses under stress, it cannot be cut/milled to high tolerances? Wow, imagine that, a material that exhibits plastic deformation and a hysteresis loop; how unique (heavy sarcasm intended).

FYI, I not only have a bachelor’s degree in Civil Engineering, I have a masters degree in Structural Engineering (earthquake forces in steel moment frames). I need no lectures about materials science, nor do I need to be reminded of the modulus of elasticity of steel, thank you very much.

If you truly believe that seeking precision is futile because wood is softer than steel, you need to pay closer attention to the world around you. Extremely high tolerances in wood are not just possible but are absolutely required in many situations. One example, and not an unusual one. In a previous reincarnation I worked for five years for a company named MetalFit (http://www.metalfit.co.jp/) making wooden structural frames from softwood glulams using fully automated CNC production lines in Japan and Minnesota. The company has changed hands several times since then, but judging from the website, the connectors and production lines appear to be the same. These robotic lines cut/milled the members using programs automatically generated from CAD structural drawings. Because the wood had to accommodate precision metal connectors and drift pins to transmit compression, tension, bending and shear, cutting and milling tolerances were +- .0008". The members were all pre-cut in the factory and assembled on the jobsite, so dimensions had to be very tight to prevent Murphy's Law of Accumulated Errors from making assembly difficult. Of course, the glulams were re-milled upon delivery to bring them up to acceptable tolerances before they went into the line.

Cutting wood to such tolerances is easy if the machinery is well designed and maintained and setup properly. Anybody in the industry knows that. For that matter, so long as its arbor is running true and the blade doesn't wobble too much, your tablesaw can do it. But believe me, only insanity can result from attempting to setup your tablesaw to cut those sort of tolerances if your square is .012” out of wack like the Starrett combo I rejected was.

The Germans do a pretty good job at precision machining of wood too. Fascinating machines. http://www.hundeggerusa.com/

But let's set half-baked engineering theory, and willful ignorance of modern commercial woodworking practice aside for now, and shift our focus back to handwork. Do you suppose that a tenon joint layed out with a sloppy square, perhaps resulting in a gap at the shoulder, is the best anyone should strive for? If you are happy with that sort of joint, blissful art thou. I prefer to use a square I know to be accurate (no, not a diemaker's square, dagnabit, pay attention), and if the shoulder turns out to be sloppy, well then that's may fault for not cutting to the line. I like Sean’s words about him disappointing his tools. No doubt we all do that to one degree or another. But just because I can’t cut by hand to the ten-thousandth of an inch my micrometer can measure, must I be satisfied with +- .012” levels of accuracy? Of course not. The golden mean lies somewhere between these two extremes, and each of us makes a decision, intentionally or no, about what that precision will be. My intention was to encourage you to ask yourself “how do I know my tools are accurate?” From that starting point, you can make an intelligent decision.

I am fine with folks that don't need or don't want their layout and measuring tools to be straight and square. I could not care less if their little projects are lop-sided. But (here’s the part where I puff up and turn red) to spout half-understood materials science principles to support the notion that precision in woodworking is futile is not to talk about the "real world," but rather about "Wayne's World."

[Kees Heiden]

I hear you about only using the super precise only for testing your other meassuring tools.

But still, you are way out with your recommended specs. That kind of precision is not neccessary, especially not in handtool woodworking, the subject of this forum.

A good square is good to have. Check it on a straight section of your bench, that's good enough. And yes, squares wear out. A;ll the old ones I find are rife for the garbage can.
A good straight edge is good to have. A couple of years ago I broke down and bought the cheapest i could find with a DIN spec, about 50 euro. It's nice to have to check plane soles. But now I like working with wooden planes and the thing becomes mood. In daily work I like a wooden straight edge, and indeed check it from time to time to the steel one.
A micrometer? No idea what that is usefull for.

I would recommend you to have a look into Peter Follansbee's work. It might change your attitude to woodworking to a more reasonable level.

[Stanley Covington]

But still, you are way out with your recommended specs. That kind of precision is not neccessary, especially not in handtool woodworking, the subject of this forum. A micrometer? No idea what that is usefull for. I would recommend you to have a look into Peter Follansbee's work. It might change your attitude to woodworking to a more reasonable level.

Kees, thank you for your comments. Please help me clearly understand the points you are trying to make.

1. You say that the specs I recommended are "way out." Good enough. But remind me where I recommended any specs in any of my posts. Looking back at my posts I see that I did mention that I found .012" runout over 12" to be excessive for me. I also wrote that a particular product I was once involved in manufacturing commercially was milled with CNC machinery to tolerances of +-.0008". But I hope I would never be so foolish as to recommend specs or tolerances to anyone on this forum. That's something for each man to decide for the individual project he is working on, either on his own, or with his client, at least in my opinion.

2. You wrote: "That kind of precision is not necessary..." OK, what "precision" do you refer to Kees? I didn't give any specific numbers, so do you mean any precision that requires machinists tools? If so, is the level of accuracy of your 50 Euro straightedge, or the "section" of your bench, the standard we should all follow? What are these standards for precision all woodworkers should follow? I respect your opinion, but what is this sweeping assertion based on? Rough work like what Peter Follansbee does? Are you suggesting that his is the only kind of work that can be done with handtools? Is that the only kind of woodworking worthy to be discussed in this forum? Must instrument makers and sashimonoshi be satisfied with these standards of precision you seem to share with Peter Follansbee?

BTW, I have a signed copy of his newest book on making a joint stool. Great book. But that isn't the whole world of woodworking.

3. I apologize if my attitude needs to be adjusted to more reasonable levels. I will reread Mr. Follansbee's book tonight and search for ways to adjust that troublesome thing. But can you help me out and tell me what you think those levels should be?

Thanks,

Stan

[Kees Heiden]

I don't want to sound too much like a knowit all. But I like the discussion. I am an engineer myself, but like to work purely with handtools at home in my gardenshed.

12 thou over 12". That's the same as 1mm over a meter isn't it? That's indeed too much for a straightedge. I have no idea how precise my cheapish DIN normed one is, but it's definitely straighter then that. When you have a reasonable straightedge, you can check the side of your workbench and plane it to straight and use that to test your squares. It's not dificult to plane a piece of wood straighter the 1mm over a meter.

I still don't understand the need for a micrometer though.

I just feel that you can do fine woodworking without the latest in Mitutoyo measuring equipement. My first project, a pine box with panneled sides was made with some "Home Depot" tools. It still looks perfectly fine with tight shoulders for the M&T joints. I used a Sandvik square, unchecked, straight form the store.

And I mentioned Peter Follansbee to show the other end of the spectrum of woodworking. His work does look rough from close up, but it looks very nice from a little distance. Pieces like that from the middleages have survived very well over the centuries, and I admire them a lot.

In fact the more I learn about woodworking, the more I steer away from the precision stuff and just try to increase my skills so I can make cuts without too much of measuring around. Feel and emotion against logic and engineering. That kind of mindset.