Decent try square

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
bugbear":1r88wzx4 said:
......Werner Heisenberg says it's impossible. There is a big conceptual gap between very small and zero......

They're wrong too.

What is so special about 90 degrees which makes it impossible to achieve? 89.99999999999999 is possible, 90.00000000001 is possible. Why not 90?

I am making a point about logic and the precision of language, not about the precision of squares. If I went out to the workshop and made 10 squares out of wood, they'd likely all be different, but there is no reason on this planet why one (or more) of them might not fluke exactly 90 degrees. (Of course, I would have no way of knowing whether I'd achieved this or not).

I fully understand the point you are trying to make: that manufacturing X number of squares and guaranteeing they'll all be 90 degrees is impossible. I agree with this: it is impossible. But that isn't actually what you said.
 
You could make your own.
Bridge City clone:

539a7f168a5007c4a8f7435427cc3a6e.jpg


http://www.bridgecitytools.com/default/ ... quare.html

Rod
 
MikeG.":aebdzi4m said:
bugbear":aebdzi4m said:
I fully understand the point you are trying to make: that manufacturing X number of squares and guaranteeing they'll all be 90 degrees is impossible. I agree with this: it is impossible. But that isn't actually what you said.

Many years ago I had a similar conversation with my professor in an advanced calculus course. The subject was converging on zero in an infinite series of halving the distance between two points. The professor, quite correctly, stated that the goal could never be reached mathematically. However, my argument was sometimes close was good enough, and provided an example. I told him that if his daughter was at one end of the room and I was at the other end of the room, and I divided the distance between us by half each time, I am quite confident he will be a grandfather without ever reaching zero.

As a teenager, I apprenticed as a machinist for one year and a cabinet maker for two years. While I appreciate the drive for precision, I value consistency and repeatability more. I use Class I tape measures and Class II folding rules because they are more predictable than the other options. I am just as picky with my squares, but am willing to accept a bit of error as long as I can correct it with a file after testing in the shop. Other times, I might not care...measure with calipers, mark with a grease pencil, and cut with an axe.
 
MikeG.":4zhspux5 said:
bugbear":4zhspux5 said:
......Werner Heisenberg says it's impossible. There is a big conceptual gap between very small and zero......

They're wrong too.

What is so special about 90 degrees which makes it impossible to achieve? 89.99999999999999 is possible, 90.00000000001 is possible. Why not 90?

I am making a point about logic and the precision of language, not about the precision of squares. If I went out to the workshop and made 10 squares out of wood, they'd likely all be different, but there is no reason on this planet why one (or more) of them might not fluke exactly 90 degrees. (Of course, I would have no way of knowing whether I'd achieved this or not).

I fully understand the point you are trying to make: that manufacturing X number of squares and guaranteeing they'll all be 90 degrees is impossible. I agree with this: it is impossible. But that isn't actually what you said.
It's actually not possible to measure a square to unlimited accuracy either, so tiny errors are both unavoidable and unmeasurable. Since the arms aren't straight to zero error, the angle between them isn't even defined to unlimited accuracy.

It's all (at a tiny scale) mushy.

None of this matters, of course. The workshop conditions needed to fully exploit squares with errors of < 1/100,000" (which can be made) are so absurd, no one would do it anyway.

BugBear
 
Stevedimebag":y5lv916g said:
Thanks for the advice folks. True enough about the starrett - I will just buy it and if it's no good I'll just send it back.
There's a fair bit of advice online about squaring up try and combi squares... would that not be better than spending on postage for a £6 square?

Brandlin":y5lv916g said:
That's 1 mm out over a length of a metre...
In engineering terms that's not accurate at all.
And 1 metre out over a kilometre, which is about how accurate our mortars were - Depends what kind of engineering you're doing and what kind of thing you're 'engineering'. Snipes would be horrified at the inaccuracy, but artillery blokes would be fairly chuffed with that kind of precision. :lol:

So at 50cm you're ½ a mil out? That's about 1/64... Surely that's "close enough for an Englishman" in woodworking terms? :p
 
Tasky":2l30gsf1 said:
So at 50cm you're ½ a mil out? That's about 1/64... Surely that's "close enough for an Englishman" in woodworking terms?

It depends.

Let's say you're making a cabinet with a 500mm wide door and you want an even gap all around the door. Let's just focus on the gap along the top edge. If the gap was 1.0mm at one end and 0.5mm at the other then, to my eye at least, that would be glaringly obvious and unnaceptable.

Alternatively say you have a 2000mm long dining table that's 2mm narrower at one end than the other, I'd hope I'd never be that far out, but if I were I doubt anyone would ever notice.
 
Brandlin":2hhbtzzd said:
That's 1 mm out over a length of a metre...
In engineering terms that's not accurate at all.
And 1 metre out over a kilometre, which is about how accurate our mortars were - Depends what kind of engineering you're doing and what kind of thing you're 'engineering'. Snipes would be horrified at the inaccuracy, but artillery blokes would be fairly chuffed with that kind of precision. :lol:

So at 50cm you're ½ a mil out? That's about 1/64... Surely that's "close enough for an Englishman" in woodworking terms? :p[/quote]

My point with the numbers was that what is seen as "acceptable" tolerances varies by discipline, application and individual perception.

As i said in my fist post. "How square is square ENOUGH?".
 
If I need something 'square enough' I use an ancient Rabone engineers square. If I need it square-ish, I'll use whatever else is nearest.
 
custard":puno20ny said:
Tasky":puno20ny said:
So at 50cm you're ½ a mil out? That's about 1/64... Surely that's "close enough for an Englishman" in woodworking terms?

It depends.

Let's say you're making a cabinet with a 500mm wide door and you want an even gap all around the door. Let's just focus on the gap along the top edge. If the gap was 1.0mm at one end and 0.5mm at the other then, to my eye at least, that would be glaringly obvious and unnaceptable.

Alternatively say you have a 2000mm long dining table that's 2mm narrower at one end than the other, I'd hope I'd never be that far out, but if I were I doubt anyone would ever notice.


but ... but .. couldn't you be out by 0.5mm on one side if the wood happened to move more on that end? .. is that something that can occur?
 
transatlantic":2tmcc6kp said:
custard":2tmcc6kp said:
Tasky":2tmcc6kp said:
So at 50cm you're ½ a mil out? That's about 1/64... Surely that's "close enough for an Englishman" in woodworking terms?

It depends.

Let's say you're making a cabinet with a 500mm wide door and you want an even gap all around the door. Let's just focus on the gap along the top edge. If the gap was 1.0mm at one end and 0.5mm at the other then, to my eye at least, that would be glaringly obvious and unnaceptable.

Alternatively say you have a 2000mm long dining table that's 2mm narrower at one end than the other, I'd hope I'd never be that far out, but if I were I doubt anyone would ever notice.


but ... but .. couldn't you be out by 0.5mm on one side if the wood happened to move more on that end? .. is that something that can occur?

Not if you've made the cabinet door correctly. The bigger problem is installing a cab in a clients home and finding their floors aren't level, which then affects the gapping. But once you've levelled it up it'll stay accurate. Here's a good example, not my work but it illustrates the point. This particular tool cabinet has been in place for several decades, it's opened and shut repeatedly, the weight of tools hanging inside the doors varies, but the accuracy of shut lines would still put a BMW to shame.

Wall-Tool-Cab.jpg


The idea that "it's just wood and wood moves" isn't exactly wrong, but it is often overstated, I guess what I'm saying is don't let that become an excuse for sloppy work.
 

Attachments

  • Wall-Tool-Cab.jpg
    Wall-Tool-Cab.jpg
    108.4 KB
transatlantic":38ruid41 said:
but ... but .. couldn't you be out by 0.5mm on one side if the wood happened to move more on that end? .. is that something that can occur?
That's what I'd be worried about - Wood swelling or shrinking.
 
MikeK":1o1azl9l said:
MikeG.":1o1azl9l said:
bugbear":1o1azl9l said:
I fully understand the point you are trying to make: that manufacturing X number of squares and guaranteeing they'll all be 90 degrees is impossible. I agree with this: it is impossible. But that isn't actually what you said.

Many years ago I had a similar conversation with my professor in an advanced calculus course. The subject was converging on zero in an infinite series of halving the distance between two points.......

That's not a direct comparison. The claim was that 90 degrees was impossible. Well, I've heard no explanation why the only impossible angle out of the infinite number of possible angles is 90 degrees. I accept that we can't measure it. I accept that even if we could measure to infinite accuracy, that manufacturing squares that were perfect every time would be impossible. I will not accept, however many different ways it is put, that there is only one angle which is impossible to achieve, and that is 90 degrees. Like I said, this is a logic and semantic argument, and not a discussion about squares.
 
Tasky":13zo6pa5 said:
That's what I'd be worried about - Wood swelling or shrinking.

Firstly you design around that, frame and panel, veneer on ply, etc. That's what woodwork is largely all about.

Secondly, take a section of a dry hardwood board, use a timber that's representative of the woods you're likely to use, hang it up and accurately measure the width across the grain. Keep measuring it and record those measurements across a full year. You might be surprised at how stable it is, at least once any initial drying has occurred.

A lot of the commentary on contemporary woodworking comes from the US, the majority of the US suffers from far bigger wood movement problems than we do by virtue of their continental climate. In the UK we have a maritime climate, it's mildly damp all year round and our temperature swings really aren't all that dramatic. I'm not saying wood movement in the UK is nothing, but I am saying it's manageable and doesn't preclude quite high levels of fit and finish for your work. I make my drawers to run very snug side to side, there's a bit more allowance up and down but side to side there's no hint of rattle. I've been doing that for many years and I can go back to examples I made twenty or thirty years ago and they've retained that piston fit snugness without any sticking problems.

Don't let wood movement become a lazy excuse for sloppy work.
 
bugbear":r8ursijf said:
MikeG.":r8ursijf said:
bugbear":r8ursijf said:
......Werner Heisenberg says it's impossible. There is a big conceptual gap between very small and zero......

They're wrong too.

What is so special about 90 degrees which makes it impossible to achieve? 89.99999999999999 is possible, 90.00000000001 is possible. Why not 90?

I am making a point about logic and the precision of language, not about the precision of squares. If I went out to the workshop and made 10 squares out of wood, they'd likely all be different, but there is no reason on this planet why one (or more) of them might not fluke exactly 90 degrees. (Of course, I would have no way of knowing whether I'd achieved this or not).

I fully understand the point you are trying to make: that manufacturing X number of squares and guaranteeing they'll all be 90 degrees is impossible. I agree with this: it is impossible. But that isn't actually what you said.
It's actually not possible to measure a square to unlimited accuracy either, so tiny errors are both unavoidable and unmeasurable. Since the arms aren't straight to zero error, the angle between them isn't even defined to unlimited accuracy.

It's all (at a tiny scale) mushy.

None of this matters, of course. The workshop conditions needed to fully exploit squares with errors of < 1/100,000" (which can be made) are so absurd, no one would do it anyway.

BugBear

Hello,

I have got the worlds only 90.00r degree square in my posession, just by lucky hap. I have it safely locked up in a box, with Schrodinger's Cat, until the day someone invents a uni-molecular pointed pencil made from Graphene, to make full use of it. Oh and a uni photon thick laser to sharpen it.

However, I would advise the OP to but a decent combi square from a reputable maker, I have a Starrett, (American made, not the Chinese made rubbish the OP refers to) others fine makes are available. It is a once only purchase and removes all doubt. Having cheap squares never ends well.

Mike.
 
woodbrains":3pkakvis said:
Hello,

I have got the worlds only 90.00r degree square in my posession, just by lucky hap. I have it safely locked up in a box, with Schrodinger's Cat, until the day someone invents a uni-molecular pointed pencil made from Graphene, to make full use of it. Oh and a uni photon thick laser to sharpen it.
Sadly, the first time you use it, the "observer effect" will destroy its perfection. :(

So, just like Bridge City tools, you have to own it, but not use it... :mrgreen:

BugBear
 
custard":eaoq5gzq said:
Don't let wood movement become a lazy excuse for sloppy work.
I was thinking the opposite, as I'd be more likely to overcompensate for it - Planing down to 1/8th of a mil for absolute precision, only to have it all thrown off and fall apart when Winter rolls around...

I know there is at least some swell in the UK, as our house doors only stick when it's cold.
 
Tasky":2ea4t5kn said:
I know there is at least some swell in the UK, as our house doors only stick when it's cold.

Same here in Hampshire. But talk to New Hampshire furniture makers and they'll tell you their doors don't just stick, they jam solid...but only in the summer. That's why I'd encourage a UK maker to base their actions on local conditions, not stuff culled from the US dominated internet.
 
MikeG.":3vmas4vr said:
bugbear":3vmas4vr said:
.........If your answer is "I want it exactly square" no one can help you. Such a thing is not, and cannot be, made........
That's simply not true.
No, it really is true. A perfect 90 degree angle exists only in mathematics. In the real universe, we can't measure things shorter than a planck length according the laws of physics as we understand them (and it's hard to measure something to less than the wavelength of the radiation you use when measuring it) and the exact position of something can't ever be known at the quantum level, all of which have normally trivial effects at the macro level but one of the effects you can be certain of is that you can't manufacture something perfectly. There will always be a tolerance, even if that tolerance is down in the parts-per-gazillion range or lower.

Mind you, this is wood. It's made of cells. They're 20-30 micrometers (I'd say microns but someone would ask me how many sixty-fourths that was) across. So the amount of accuracy you need really can't get lower than that, the medium doesn't support finer granularity. And most of our tools work in larger increments - I could probably fettle my best handplane to take a shaving I could see through if I worked hard enough at it, which would be somewhere below 0.1mm. So if your square is accurate to 0.05mm over its range (say, 30cm), that's about one part in 6000, or 0.016%.

Now that's readily achievable even with modest engineering standards these days.

So I just went to Proops and bought some of their squares and they've been more accurate than me ever since. My cheapo combination square, mind you, was trash so I binned it until I could justify buying one of those nice Moore&Wright six-inch sliding squares (they're handy for checking mortices for straight edges).

TL;DR: We can't measure things to a planck length, but that's okay, because we're more interested in accurate plank length.

boom-boom.jpg
 

Latest posts

Back
Top