Proofing a square

[mklotz]

A recent thread about proofing squares...

http://www.homemodelenginemachinist.com/index.php?topic=4094.0

got me to thinking. How could one quantify the accuracy of his square using only items available in the average shop? (Testing a square by drawing lines with two different orientations of the square will show whether it's more or less square but won't quantify how far out it is.)

I came up with the scheme shown in the attached sketch. The square is placed blade up on a flat surface. A base is fitted with a mast that holds a spacer ring near the bottom of the mast and a DTI near the top. This base is placed such that the spacer ring touches the square's blade. In this position the DTI is adjusted to about half deflection and zeroed.

Now the base is slid over the beam of the square (a cutout in the base allows this to be done) and the spacer ring is pushed in contact with the opposite side of the beam. The reading on the DTI now indicates how far out of square the square is.

What do you think? Would it work?

 

[Maryak]

Marv,

I think it would work assuming that the blade of the square is parallel.

Best Regards
Bob

 

[mklotz]

I think it would work assuming that the blade of the square is parallel.

Yes, I should have stipulated that one mikes the blade first. I always do that when I get a new one, so I forgot to mention it.

 

[Noitoen]

Assuming that the mill's table x,y is square, then fixing the square to the table, with the dial you can check the squareness of the square. ;D ;D ;D



Helder

 

[itowbig]

humm interesting. it makes me think a bit more now.

 

[SmoggyTurnip]

Very interesting. Made me think ... suppose you used the setup as metioned for step 1 and then for step 2 instead of using the other edge of the square you placed a flat surface against the edge you just checked (a 1-2-3 block or parallel) then checked against the back side of that. This way it would not matter if the back side was parallel and you could also check squares that are triangular shaped

 

[mklotz]

Smoggy,

How about a sketch of what you have in mind? What I'm envisioning from your description doesn't make any sense to me.

 

[tel]

Now this might seem a somewhat simplistic view, but in my opinion a square has to be just that - square. If it's out the degree out outness doesn't really matter, you either throw it away, or deal with the error. To deal with the error you just need to set the stock up truly square to the machine and then mill/grind/whatever the blade square to the stock.

 

[mklotz]

Now this might seem a somewhat simplistic view, but in my opinion a square has to be just that - square. If it's out the degree out outness doesn't really matter, you either throw it away, or deal with the error. To deal with the error you just need to set the stock up truly square to the machine and then mill/grind/whatever the blade square to the stock.

And, after having done this mill/grind/whatever, how do you prove that the square is now square?

 

[SmoggyTurnip]

Does this make sense?

 

[Davidh]

Draw a square line and turn the sqaure over and check.



If it is no good then stick it on ebay.....

 

[putputman]

When I was in trade school, back in the 50's, one of our projects was to make a "cylindrical square". I have used it on very few occasions since I made it, but I know it is absolutely square each time I do use it.

Mine happens to be hardened and ground. I see no reason why anyone could not make one on any lathe. If the diameter is the same from end to end, (this is extremely important) and you face off the end, it has to be square.

Mine is about 2 1/2" diameter X 3" high.

You can use this to check your instruments or set ups.

 

[itowbig]

now theres something i did not think about.
thank you putput. i new this was a great place to be lots of brain power and willing to share those ideas. thank you

 

[kvom]

Using 2 squares, put one against the other on a flat surface. If there's no gap then they're either both square or out the same amount. Then reverse them so that the other legs are touching. If no gap then both are square.

 

[tel]

And, after having done this mill/grind/whatever, how do you prove that the square is now square?

Enter it in a square dancing competition, if it wins ........ ;D

 

[itowbig]

:big: tel lol :big: :bow:

 

[gilessim]

I'm surprised Marv, mathematician that you are, that you didn't bring Pythagorus in to the equation here! but, I guess you'd need a pretty big mic. to measure across the corners!

Giles

BTW I have a 6" square that my granddad made in 1911 from old files (it has his name engraved on it with the date!)that is still as good as when he made it, it's been in my tool bag for 25 years or so and I promised him that I would use it!

 

[PhillyVa]

Giles, Good for you...I would like to see that square your granddad made. Can you post a pic please?

BTW I have a 6" square that my granddad made in 1911 from old files (it has his name engraved on it with the date!)that is still as good as when he made it, it's been in my tool bag for 25 years or so and I promised him that I would use it!

Regards

Philly

 

[Bill S]

One thing that hasn't been mentioned is what tolerances squares are made to. How square is square? I like the cylindrical square for testing though, practical, doesn't need much of a setup and you could use a feeler gage to see how far a bad square is out. Which raises another question. How would you fix it if it were out. I have trued combination squares to a 64th in 10" for construction use, but what about a rigid square to a thou or less?

 

[tel]

I may be a square, but I've been around.