Let me know what you guys think.
I've been working on a new method (new?) for checking the squareness of a square using a dial indicator. The method works in theory. I've tried it and it seems to work in practice. A caveat is that the square needs a thick edge to support a stylus.
I am going to work on a video demonstrating the process.
Seems totally unnecessary. A pencil and a straight edge (a la sheet good cutoff) is all you need to check if it's actually square. And it's fast to set up.
read the article.
A minor caveat to this method is that it can be difficult to discern sm= all gaps between the two pencil lines (especially with a thick pencil lead)= . The most you might be able to detect is a 0.010" difference which equates= to a minimum detection of 0.036 degrees with an 8" square. Another caveat is that the edge you place your square against must be p= erfectly flat, otherwise you will not get an accurate calculation of your s= quare's angle error.
The dial indicator method is 10X more accurate.
square using a dial indicator. The method works in theory. I've tried it and it seems to work in practice. A caveat is that the square needs a thick edge to support a stylus.
Sinced/dx (arctan) is close to 1 near 0, yourerror is not very big, but your equation which divides an angle by L (before applying arctan) doesn't make sense. For instance, it suggests that if you choose L great enough that your error will be as small as you need it to be. I did not try to redo your derivations, but I am willing to do so if we don't find a concensus. By the way,you might sketch a triangle somewhere(tan = opp/adj)for those that may be a bit rusty at doing trig.
Bill
I appreciate the point that you are trying to make, but I don't find the argument given to be as rigorous as it should be.
I'd enjoy seeing a re-write. I'll help here if you like. I don't think that L was "well-defined" (consistent) in the present argument Bill
On 1/3/2013 6:21 PM, snipped-for-privacy@garagewoodworks.com wrote: ...
What difference does it make what the numeric value is?
If it's enough to detect that way, it's enough that for really precise layouts it's off.
So, what you gonna' do at that point, anyway even if you do know _precisely_ how much that is? Eggs-ackly the same thing as if you only know it's off---either ignore it and go on, go get a new Starrett, or adjust it until it does pass (the peen method works a charm for that).
If it passes the pencil test, it's plenty close enough for woodworking; even fine woodworking. The wood itself moves more than that.
--
square using a dial indicator. The method works in theory. I've tried it and it seems to work in practice. A caveat is that the square needs a thick edge to support a stylus.
Trust me, the math works. Draw a few squares that are off by a little in CAD and check.
Use my online calculator:
If you don't know how much it's off you can't make a call either way.
So why bother checking?
What is a "pass" on the pencil test?
L is the length of the square's edge. It was defined adequately.
square using a dial indicator. The method works in theory. I've tried it and it seems to work in practice. A caveat is that the square needs a thick edge to support a stylus.
If you are going to present the math, then do so (properly).
The sentence tht you wrote: "The answer given will be the amount in degrees that your square is off. " is terrible. Squares are not "off", degrees are not measure in amounts, and no answer was given (and no question was asked)!!
If you are going to get rigorous, then do so--let's not pawn it off. "Trust Me, the math works", should not follow your (lack of a) a successful derivation. Lets not trip to slip the argument by people just because you figure you've already worked hard enough on it! I enjoy well-done mathematics as much as anybody, surely more than some. Hope to see you try again.
Bill
square using a dial indicator. The method works in theory. I've tried it and it seems to work in practice. A caveat is that the square needs a thick edge to support a stylus.
For the same angle error, as L get longer so does abs(Delta1 - Delta2).
of a square using a dial indicator. The method works in theory. I've trie= d it and it seems to work in practice. A caveat is that the square needs a= thick edge to support a stylus.
Squares can be "off" in the sense that they are not a true 90 degrees. Eve= ry square will be "off" to some degree. Perfection is imaginary. :)
Reposted for your convenience: "For the same angle error, as L get longer so does abs(Delta1 - Delta2). "
There is no need to "try again" and I appreciate your veiled arrogance.
Where are you having trouble with the math?
Did.
From what I've seen, Dr. Grella has a pretty good grasp of mathematics and scientific principles. ;)
I challenge anyone to find a flow in my math. :)
Good luck.
Flaw Not flow. /that was intentional (grin)/
What are we building here, bookcases or space shuttles? 10X more accurate is not necessary for woodworking.
Then who is being arrogant?
and I appreciate your veiled > arrogance.
Your identifiers are poorly chosen making your "work" difficult to read. From your words/diagram, it looks like delta1, delta2, and L are all the same. I don't see the 5-degrees you mentioned (in your diagram).
Doing math is like doing design: Do-over, and do-over, and do-over. One should get quite humble about the process!
Bill
Appears to be a matter of resolution. I'm of the camp that if you don't have perfection as a goal for every step of the process, you will never scratch that itch some of us are cursed with.
At the same time, I'm just as guilty as anyone of occasionally settling for less ... mainly due to time, circumstances, the medium (woodworking), and the need to 'git r' done'. ;)
That said, you are only as accurate as your tools allow ...
I forgot to mention, what you have labeled as "References" are not references at all.
Please re-do.
Bill
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.