When a Square is not a Square...

Recently I was commenting to Ralph, that most Try-Square are only meant to be garanteed square on the inside, not on the outside. That of course raised some eyebrows. What ? That cannot be, will still need to check the inside of carcasses!?

Ah, yes, but what is square and how do we know it is??
And dont forget, we are talking about woodworking here, not machinist.

If you own a house, it's probably been a long time you saw anything square, or level :-)
But it probably started with most everything level and square, things move, compression, expansion, settling etc.

Measuring devices are no different. Hence the choice of material is important where it count for accuracy.

Square is defined as making a perfect 90 degrees between two surfaces. How can we detect or sense it? Having a known “square” angle to compare to would be one way. Looking for light between both.

The human brain has the ability to detect or see when things are approaching 90 degrees.
When you are crosscutting with a handsaw, the reflection in the saw plate let you see when you are at 90 degrees. When boring with a brace or handdrill, you can also train yourself to detect being plumb.
Similarly our senses let us get a feel for when things are level or flat.
Running your fingers over a board surface you can detect small hollows and humps. And when using a pair of winding stick for example, we can easily detect an out of true condition (twist).

All that to say, we fret too much about measuring things to be square.
The only true test is in the pudding: Can you place two piece side by side touching at the ends, and is it still flat? If so flip one piece if still flat, we have perfect 90... Regardless of what your squares are saying...

ANATOMY OF A TRY-SQUARE

The blade has its two sides parallel to each other. That is an important point to remember before you start filing away on one side of the blade to correct an out of square-Square. Being parallel, we can ensure the two inside angles will remain true to each other.
The two outside angles on the other hand relies on the stock being true and square, something impossible to garantee in this design.

The two inside angles rely on the brass plate being set (pinned) at 90 degrees to the inside of the blade (Red), and the other angle (Blue) can only be at 90 if the blade is parallel. The bottom of the stock (opposite the brass plate) cannot be relied upon to stay true

The stock being pinned rigidly by up to 4 pins some times, it is not going to be able to expand much. The other end of the stock is free to expand and contract unimpaired, resulting in a slightly taper side. How much are we talking about? That would varies with the wood species coefficient of expansion and how long is the stock. Coincidently that is why Rosewood has long been favored, it has a very small expansion ratio. What about the other side, on the inside?
That is why we used a brass "wear plate", not so much to prevent wear, but to ensure a long lasting flat reference surface, wood expansion would be pushing away from it on the unplated side.

Not really a problem since this traditional Try-Square design was never really intended to be used on the outside surface of the stock. There are better tools for that.

Ever wondered at those fancy brass inlays surrounding the pins on a Try-Square?
Watch Roy using a passer drill, a reproduction of an original set in uses at Marples UK for years.

These two are Disston No 1 Try squares.

The smaller size having three pins, the bigger one, four

These are more than decorative, they help secured the pins rigidly at what is arguably the weakest point in the square design, how to ensure these two parts, the stock and the blade, are fixed securely at 90 degrees. The brass "wear plate" is pinned to the blade in the better design to ensure a rigid 90 connection.

Or you can go to great lenght to make your design micro adjustable to make and correct an out of true square. Behold, the Clenton square

These little "problems" caused by wood expansion (who knew, wood moves? NO ones knew that until now, but ... end of polite Canadian political rant :-) are unavoidable in this design. The only way out, is to use a metal frame and put in wood inlays or used a different material, such as, in a all metal construction.
That is why, for the ultimate precision you are better off using Machinist squares, all metal construction, usually carbon steel.

Groz, Made in India. Grade B, 

inexpensive and plenty good

They are obviously only as good as the precision machining and cost accordingly. The good news is, since we are woodworkers and not machinist, we can get by using less precise Class B toolroom tools, specifically BS939 (British Standard) Grade B.
The specs are less than 0.001 inch of deviation per inch over the full lenght of the blade. Plenty good for woodworking!!!
Only downside of these precision squares is that being high carbon steel, they rust easily, some caused by our fingers oil and moisture, sigh!!
Careful cleaning it, especially when removing rust, you dont want to destroy its precision

Something else to look for is excessive wear on the blade tongue, caused by the repetitive uses of a marking knife edge against it.
Often, that wear would only be confined to a small portion of the blade, since depending on how big the square is to the job at hand, only the first few inches would be used...
That type of wear would render the blade out of true and change its critical parallel dimensions.

When you have an out of true square, if the blade is straight and parallel, i'll touch the stock to correct instead.

There were obviously wooden square prior to the metal versions, and they are a good project to undertake. They are favoured by some because of their light weight. They can be made plenty square for our needs.
Ironically wooden Try-Square are easier to maintain for true 90 on the outside edges, meaning they are more handy to reliably check the inside of carcasse.
Of course, regardless of design any square should be tested once in a while to ensure continued accuracy.
Here is a good tutorial on checking and correcting out of square square...

Another great source of inexpensive precision square are the drafting shop.

Buy good ones at specialized shops

So back to the original objection of not being able to trust the outside edge, how do we measure inside carcasse for example? Short answer is we don't.
We check squareness by checking if both diagonals are equal, if so: Square
But, yes, by using a machinist square we could check all four corners and, it may or not, be still square.

Or another style such as the English layout square would be more appropriate in this instance. This design obviously does not allow inside measurement, for which the Try-Square exist

ATC square or English layout square

Pic from Chris Schwarz Blog

See article here

This English layout square can easily be used as a Libella, by the addition of a plumb bob and a string held at the apex. With it you can check for both plumb and square.

And here are some ancient variation of the square that became lost thru the years

Yes, there exist a large variety of squares for checking squareness in various ways, approach and etc,

And for everything else, there is always the Magic Square from fellow Canuck, Chris Wong :-)

All that to say, just because you are armed with a "square" unless you know it is actually accurate, it is anything but a square... Regardless of make, pattern, condition, price point etc. It does behove us to take care and protecting them.

I leave you with Paul sellers takes on it..."being square"

Bob, chasing Rudy around the room to get my square back from him. Will need to check it after :-)