What Do All Those Zeros Really Mean?

Cactus Farmer

Active User
Registered
Joined
Sep 24, 2013
Messages
397
I'm having some internal unease with some folks use of the 4th and 5th digit when talking about machining parts. Most folks can't measure to the 0.0000 with precision, and surely not to the 5th place. By including this place holder you are implying that you can actually measure that accurately . Most can't. I have a comparator and it measures to the 1/10,000 and I can infer to the one hundred thousandth. But I would be very cautious about writing it as such. Do they not understand that to 0.000 is pretty close to perfect. I have pin gauges in minus sizes and a few in the plus range. With a mike you are hard pressed to "see" a difference. The comparator will allow me to "see" it but only with the shop warmed up to a "standard" 70 plus or minus degrees. And then only it has had time to make the machine reach the same temp. The base of the comparator is BIG cast iron and warms very slowly. Is there a need for a lesson of accuracy? Most folks know "ain't" is not proper but "irregardless" is not too. Both are double negatives......Flammable and inflammable? Is one capable of burning and the other not? No, both will toast your behind if handled carelessly.
It just irks me no end for the language to be used poorly. I want to be accurate in my making of things but I want to convey what I did as concisely as possible with a common understanding as to
what I mean with a written explanation.
It seems like the most measurements that bug me are bore diameters of barrels. And sizes of the bullets that are fired through them.
To restate the idea, 0.10 implies accurate to 1/100 of an inch. 0.100 is accurate to 1/1000 of said inch , and 0.10000 is accurate to 1/10,000 if that inch. As I stated above, my comparator has a 2 1/2 inch thimble to to move the stage but it is still 0.025 per revolution just like a standard micrometer. That spreads the 1/1000 marks to about 1/2 inch apart. There are small marks between the 1/1000 marks that corespond to 1/10,000 if an inch. If I need a quick check on a tap for size and TPI it is not a big deal to measure the needed dimention. If "class of fit" for that same tap is required I need to wait for the shop to stablize temperature wise before attempting a measurement. Temperature makes things move!
So, adding zeros after the first three places is you impling have measured something with this degree of precision.
Most jobs require 0.001 accuracy or so. Putting that extra digit after the one thousants place is not adding precision if you can't measure it with sureity. If I need a tight fit on ,say, a bearing fit, -0.002 will surfice.
0.125 is an eight of an inch. Now measure 0.124 or 0.126 with that steel rule and tell me the difference. You are better than me if you can see that little change. The same applys to micrometers. 0.0011 and 0.0012 look a whole lot alike in your warm hand.
What I am trying to convey is adding a 4th digit may look cool as everything, your work will not be any better that your measurements.
 
In a professional setting, I would agree with you.

But for my own personal use, I use four decimal places, because my DRO's are capable of four decimal places resolution and because it eliminates some of the ambiguity in the conversion of millimeters to inches and common fractions to decimal fractions. But I realize the capabilities that I and my machines have and I know what fit and finish is required to make a serviceable part. I also over-define features by multiple dimensions. I do this for my convenience even though it is a drafting transgression to do so.

I have seen instances where metric measurement was rounded to three decimal place inch measurement and the rounding error compounded over a number of dimensions in a chain to the point where the machined part was unusable. (12 iterations of holes spaced on 9 mm centers; rounding down to .354" produce a .004" offset). This can avoided to a certain extent by proper use of ordinate dimensioning. But I realized the capabilities that I and my machines have and I know what fit and finish is required to make a serviceable part.

The important thing is to properly, unambiguously, and efficiently convey the design intent. "Ain't" may not be proper use of the English language but there are far worse transgressions in our modern day communication that need to be fixed first.
 
for me I have no illusions that I can consistently hit 0.001 accuracy. I am not a machinist or an engineer but I thought that accuracy was defined separate of the target dimension such as 2.0 Plus or Minus .0001 . On my Lathe with a DRO my drawing often come out with 3 or 4 places after the decimal. This is because the number of places to print is a default setting that I don't see the need to adjust. Just the same, If I am on the Lathe turning to 1.5" diameter, I am going to turn it down until the DRO says 1.5000 Not because I believe that I will be that accurate but because I think I will be as accurate as I can and most likely better than if I stopped with the DRO read 1.502.
 
In a machine shop, if you propose to have a part made with a bunch of 5 place and more 0’s, you are going to pay dearly for that part. Holding that kind of tolerance in a machine shop is not easy and it will cost you. And IMHO a DRO with 5 places is kinda silly, especially on a manual milling machine. You might as well tape over that last digit…Good Luck, Dave.
 
Leave it to a Texan to get to the point in a round-about- way. :applause 2:Cactus (even sounds Texan...I like it:D) has made a good point. My micrometer and vernier caliper can read tenth of a thousandths (1/10000") but how sure am I of their accuracy in doing so...don't know. Also I too believe the local temperature can affect them adversely. Though the engineers at NASA supposedly work in hundreths of a thousandth (1/100000) or millionths (1/1000000) so somebody knows how to go there. Not me, I'm satisfied with + - .001 on my work. What are we trying to do? Rocket science?:grin:
 
Leave it to a Texan to get to the point in a round-about- way. :applause 2:Cactus (even sounds Texan...I like it:D) has made a good point. My micrometer and vernier caliper can read tenth of a thousandths (1/10000") but how sure am I of their accuracy in doing so...don't know. Also I too believe the local temperature can affect them adversely. Though the engineers at NASA supposedly work in hundreths of a thousandth (1/100000) or millionths (1/1000000) so somebody knows how to go there. Not me, I'm satisfied with + - .001 on my work. What are we trying to do? Rocket science?:grin:
We need to be careful to distinguish between repeatability and accuracy. You probably can make two parts that match to within a few tenths.
 
Mr. Hasler, sir,
I have never claimed to be an english major. I AM a geologist by profession and a gunsmith by avocation. I wrote this as a pet peave. I know my limitations, (as Mr. Eastwood said). 0.000 is way close. Now with scraping you can, with three surfaces and alternating them, approach nil, the virtual penaplane. Flat an the preverbial flitter.Left to erosion alone,the earth will eventually be flat, but due to the movement of the earths plates and a few pesky volcanos it will very likely never happen. I just want things to work smoothly. And if putting a butt load of zeros behind any actually acheaveable measurement, and it toots your horn plus makes your projects work any better, go for it. As for me and mine, I'll enjoy the 75 degree day in shorts and a t-shirt. I hope you have the day your deserve.........
 
Back
Top