Shaft Machining - How do I convince Vendors that the Quality is No Good? Standards? - Page 3

# Thread: Shaft Machining - How do I convince Vendors that the Quality is No Good? Standards?

1. Stainless
Join Date
Dec 2010
Location
Mifflintown, PA 17059
Posts
1,773
Post Thanks / Like
Likes (Given)
20
202
DIMENSIONING AND TOLERANCINGcimal Inch Dimensioning. The follow- ing shall be observed where specifying decimal inch dimensions on drawings: (a) A zero is not used before the decimal point for values less than one inch. (b) A dimension is expressed to the same number of decimal places as its tolerance.

As per the link provided! Need I say more ? I might add a thank you for the link!!

2. Hot Rolled
Join Date
Jul 2010
Location
corvallis,or
Posts
872
Post Thanks / Like
Likes (Given)
89
366
Originally Posted by Froneck
DIMENSIONING AND TOLERANCINGcimal Inch Dimensioning. The follow- ing shall be observed where specifying decimal inch dimensions on drawings: (a) A zero is not used before the decimal point for values less than one inch. (b) A dimension is expressed to the same number of decimal places as its tolerance.

As per the link provided! Need I say more ?
Which means exactly what it says and nothing else.

12.000 +- .005 is correct. 12.0 +-.005 is incorrect. It says nothing about significant digits or how to interpret tolerances.

You can disagree all you want but the way it is taught in schools and the way that just about everybody else will interpret tolerances is different from you. If you want to be in a position of arguing a part into tolerance based an an unusual interpretation of the specification, that is your choice.

3. Stainless
Join Date
Dec 2010
Location
Mifflintown, PA 17059
Posts
1,773
Post Thanks / Like
Likes (Given)
20
202
What it states is that any additional measurements beyond the decimals on the drawings are not to be measured! I don't know what schools you refer to but Rutgers, Stanford and MIT have proven my point! AutoCad will dimension the same way. And the many drawings I have made have never been in question! I hope you are not referring to a Vocational school! I had a rather poor machinist that wanted a raise but wasn't worth what he was being paid! I said sorry NO! He didn't mind, he had a job offering, A Vocational school teacher. Once a week he would drop in to get answers to problems he was having!!!!! Not with the students but with machining problems like Knurling!

4. Cast Iron
Join Date
Sep 2008
Country
State/Province
Nova Scotia
Posts
293
Post Thanks / Like
Likes (Given)
19
124
Originally Posted by mTeryk
ASME Y14-5M-2004 Section 2.4 Interpretation of limits

2.4 INTERPRETATION OF LIMITS
All limits are absolute. Dimensional limits, regardless of the number of decimal places, are used as if they were continued with zeros.

EXAMPLES:
12.2 means 12.20...0
12.0 means 12.00...0
12.01 means 12.010...0

To determine conformance within limits, the measured value is compared directly with the specified value and any deviation outside the specified limiting value signifies nonconformance with the limits.
Thank you. It's kind of funny that it made it this long without anyone quoting a standard. All the discussion seems worthless without referencing the standard you're using to interpret the print.

5. Stainless
Join Date
Dec 2010
Location
Mifflintown, PA 17059
Posts
1,773
Post Thanks / Like
Likes (Given)
20
202
Dimension-Ing-and-Tolerancing.pdf

(b) A dimension is expressed to the same number of decimal places as its tolerance.

I can find many references! A few I have posted the actual verbiage.
Simply put there is quite a bit of information on this topic available but not easy to post here! I know what the standards are but if you want to find it search for it! If for any other reason than to prove me wrong! However the above is one that I have located! But think about it! A dimension in thousandths is oversize by a millionth of an inch! Does that sound rationale? If you are a machinist it's easy to understand! More so a Tool and Die man will see it far more easier. The hobbyist hasn't a clue!

6. Diamond
Join Date
Jun 2004
Location
Melbourne Australia
Posts
5,458
Post Thanks / Like
Originally Posted by Froneck

7. Aluminum
Join Date
Jan 2009
Location
Maine, USA
Posts
202
Post Thanks / Like
Likes (Given)
97
90
Froneck, your reference in post #45 is simply telling the drafts person how to create a proper drawing. This ASME document is about creating proper drawings, not parts, as stated in section 1.1. This is what mTeryk is trying to point out in post #42. It says nothing about measurements as you state in post #43. In fact this entire document never discusses taking measurements or the tools to do so. It also doesn't talk about significant digits (or numbers) at all. Section 2.4 has been already quoted and you refuse to listen or understand, but it's clear:

"To determine conformance within limits, the measured value is compared directly with the specified value and any deviation outside the specified limiting value signifies nonconformance with the limits. "

The point is that by the point the drawing is created, all the significant digits stuff is taken into account and the builder/machinist/etc. just has to build within the limits, he doesn't have to figure it out. It's up to the builder to have the tools to fall within the limits.

If you continue to argue this against all of us trying to explain it, we will do as your professors have done and drop it. This is my last post in this thread, you can lead a horse to water.....

8. This is fascinating. I do see the point about tolerance. Yet I have always been subject to the wishes of the customer. It is always a pain to run something so close and not use a gage to test fit and function here. My view is the part is in and the problem is too nit picky. As a vendor with that much trouble that unless the issue is easily fixed and current parts bought off I would increase the price. GD &T is supposed to in concept make this understandable.

OTOH the customer has the justification to do as they wish yet a callout is required to make that clear. For the future.

Many at this point will never change a situation like this especially if they have a practice of having issues like this come up. The end user will in fact use those shafts if they need them and do not have the time to burn the vendor. LetÂ’s admit that this seems petty and that those parts are good this time. Provide a gage if you want to hard check fit and function.

As is the end user user needs to tighten up the callout or add zeros or what is needed to get a good part. Between QC and the Machinist the Machinist understands better the fit and function thing better if they have ever made various fits and had th be sure one to another is fitting right. This can not be done here because one component is in question and the other component is not there.

I have seen customers use this kind of thing to lower the amount they pay a vendor because it is not right. They develop a reputation and are avoided like the plague. Such actions intentional or structural are a problem especially most for the shop who is under the gun when their part is in tolerance and they need to be paid.

9. Stainless
Join Date
Dec 2010
Location
Mifflintown, PA 17059
Posts
1,773
Post Thanks / Like
Likes (Given)
20
202
Originally Posted by mTeryk
ASME Y14-5M-2004 Section 2.4 Interpretation of limits

2.4 INTERPRETATION OF LIMITS
All limits are absolute. Dimensional limits, regardless of the number of decimal places, are used as if they were continued with zeros.

EXAMPLES:
12.2 means 12.20...0
12.0 means 12.00...0
12.01 means 12.010...0

To determine conformance within limits, the measured value is compared directly with the specified value and any deviation outside the specified limiting value signifies nonconformance with the limits.
In searching for the site for the text I posted that came up file not found I did find the site that listed the above examples! However in the same site reading down it states

"Just to set the record straight, you are only talking about metric dimensions.

Metric dimensions do not use trailing zeros to hold decimal places for tolerancing, English dimensions do. This has caused us a lot of problems with the default tolerance block on our drawings. There is an exception when using bi-lateral or limit tolerancing.

Metric dimensions less than 1, use a leading zero, English dimensions do not."

That is because in metric there is qualifying notation, like km, m,cm,mm and so on. Just like a "foot or feet" is 12.0000,,000 inches! There is no need for trailing zeros! But 2 feet does not indicated 24.000000 inches, 2. feet will but only at least 24 inches. If 2.0 feet it will be at least 24.0 inches. That has to do with certian or uncertian significant figures. Similar to adding significant figures. 2.1 + 1.01 and 3.111 and .0101 = 6.2 Not 6.2311

Check gauge pins, a .100 pin is not .100000", a .100 gauge block is not .100000000" Call Starrett or any gauge maker and ask for a 1.000 hole gauge and a 1.00000 hole gauge, I'll bet the 1.00000 gauge price is a lot more than the 1.000!

10. Stainless
Join Date
Dec 2010
Location
Mifflintown, PA 17059
Posts
1,773
Post Thanks / Like
Likes (Given)
20
202
Life Sciences Cyberbridge Is something I found that somewhat explains significant numbers determine the measurement. Though in metric it will apply to inch too!
Check out the listing but here is a small portion.
"If you were to make your own measurements, your significant digits should include all of the measurable digits (the digits that correspond to the marks on the ruler) as well as one estimated position beyond the smallest measureable digit (the 5 in 3.5 cm, and the 2 in 3.52 cm). Likewise, if you read that someone else has found a leaf that measures 4.568 cm, you can assume that this person measured with a ruler that had markings every hundredth of a cm and estimated the final digit (the 8 in 4.568)."

11. Cast Iron
Join Date
Nov 2018
Country
UNITED STATES
State/Province
Michigan
Posts
319
Post Thanks / Like
Likes (Given)
159
180
Originally Posted by Froneck
Check gauge pins, a .100 pin is not .100000", a .100 gauge block is not .100000000" Call Starrett or any gauge maker and ask for a 1.000 hole gauge and a 1.00000 hole gauge, I'll bet the 1.00000 gauge price is a lot more than the 1.000!
Yeah, because they have different tolerances. That's nothing to do with significant figures, other than those which need to be specified to define the limit. The limit itself does not depend on the number used to describe it. 1.000 ± 0.005" is the same as 1.0000 ± 0.0050" in every way other than the amount of ink used to print it.

12. Diamond
Join Date
Oct 2005
Country
UNITED STATES
State/Province
Pennsylvania
Posts
15,633
Post Thanks / Like
Originally Posted by BoxcarPete
Yeah, because they have different tolerances. That's nothing to do with significant figures, other than those which need to be specified to define the limit. The limit itself does not depend on the number used to describe it. 1.000 ± 0.005" is the same as 1.0000 ± 0.0050" in every way other than the amount of ink used to print it.
Sounds like people are asking "What is the tolerance of my tolerance ?"

13. Aluminum
Join Date
Jan 2009
Location
Maine, USA
Posts
202
Post Thanks / Like
Likes (Given)
97
90
Originally Posted by digger doug
Sounds like people are asking "What is the tolerance of my tolerance ?"
Exactly one person...

Froneck your leaf example is correct but you aren't applying it correctly, you're going at it backwards. The tolerances/limits on the drawing are the result of any stackup of such significant digit errors that might exist throughout the calculation and design process. As Y14.5 states, the limits are absolute. They are the end of the line. Not the beginning. The designer is not required to first find out what the machinist has for measuring tools, the machinist must use the tools necessary to ensure he meets the design. Documents like Y14.5 are there to foster good understanding between those two parties universally.

14. Stainless
Join Date
Dec 2010
Location
Mifflintown, PA 17059
Posts
1,773
Post Thanks / Like
Likes (Given)
20
202
The dimensions on a drawing determine the accuracy needed. To simply add zeros to every number can't be used. Going the other way if 12.2" is on the drawing. Maybe something simple like and electrical box and +/-.2 tolerance is used then a box that is put on a CMM measures 12.400001 is no good? But one that is 12.400000 acceptable! It has to do with significant numbers. The maker of the drawing indicates the level of measurement buy the number of decimals used not only zeros! The dimension in this case states a box larger than 12.4 is not acceptable, therefore a box 12.5 is rejected. Furthermore if that tolerance was used I would quality control would use a machinist scale, heck probably a Stanley tape measure! They would probably reject anything over 12-7/16"
Same with gauge pins .100 minus ZZ tolerance is -.0002 but I'll bet that one measuring .9799999 is not rejected and included is a set. Also is a gauge pin marked as to actual size when in use? What about a Micrometer that has been calibrated but as most micrometers they too have an accuracy tolerance, an item with a +/-.005" tolerance is measured .005" over the stated size is it good? Are standard .XXX micrometers calibrated to .00001? What about the actually accuracy tolerance of the micrometer? Significant numbering states the need to use a measuring device capable of thousandths!

15. It sounds like a refresher course is needed on the meaning of the terms "equal", "greater than" and "less than". Digits have nothing to do with it.

16. Aluminum
Join Date
Jan 2009
Location
Maine, USA
Posts
202
Post Thanks / Like
Likes (Given)
97
90
I'm not sure why but here goes.

Froneck, if I understand you, you're saying that if something rounds down to 2.4 instead of up to 2.5 it's acceptable if 2.4 is the upper limit on the drawing.

If that were the case what if something actually measured 2.4499. You've got your rule reading in 0.1 increments because that's all the drawing requires and you're allowed to estimate the last digit. Low and behold it lands right in the middle. What do you do? Get out the 0.01 rule? Now it lands right on a line. But you've now measured it twice over 2.4. How far do you go to spit hairs between 2.4499 and 2.4501 when the drawing only has a 0.2 tolerance and you know you're above 2.4.

The answer is you don't, because 2.40..0 is the upper limit. You can measure that with the first tool you had out as intended unless it happens to land right on the line, in which case you shouldn't bet your paycheck it's actually a good part.

17. Diamond
Join Date
Jan 2007
Location
Flushing/Flint, Michigan
Posts
7,976
Post Thanks / Like
Likes (Given)
403
6596
Dimension "1/2 inch" or 0.5 and a tolerance of +/.-0002.
That is the call out for a SNA-432 insert. What does one do?
What is allowed for the max IC dimension?
Can you cheat those walls by 50 millionths, a tenth, five tenths, ten thou.?
I get the significant digits thing and as a engineering student would have bought in lock, stock and barrel.
Latter I would enter the world of QC and measuriment uncertainly and found my earlier world crushed.
Bob

18. specify to have the journals ground.

but its going cost you......

19. Diamond
Join Date
Jan 2007
Location
Flushing/Flint, Michigan
Posts
7,976
Post Thanks / Like
Likes (Given)
403
6596
specify to have the journals ground.

but its going cost you......
Why force a process you see as best onto a vendor?
People hard turn bearing races way inside this.
Grind is nice but I would never spec it.
If you can mill or turn inside with nice machines that's all fine with me and it's sort of crazy what these once crude machines can now do.
The step is a problem ... maybe. Or maybe not.
It certainly is not clean and won't feel good like pissing down your leg so one can "sandpaper" it out to get the yes clean feeling which does not work in function but looks nice.
And here the inspector with his degree from MIT or maybe less makes the decision on buy or take the vendor out back of the woodshed.
I've been to those trips.
Bob

20. There might be some confusion as Froneck(?) is stating significant digits vs tolerance.... (maybe?)

I was taught by my 'elders'* significant digits imply tolerance, kind of.

4 place decimal = +/-.0005
3 place decimal = +/-.001
2 place decimal = +/-.01 (see 2 zeroes instead of 3 here)
1 place decimal = +/-.1

This should be in the title block with a note - unless otherwise stated- but we all don't work off fancy smancy ASMEY.147589/ANSI12 prints.

In this respect, I agree with him, I see +/-.005, I think "oh don't mean much, caliper it..." which would result (in my example of .0001" out) in "it's plus .005, high limit, but it is ok"

My stake in this whole thing was, I was the one that had to remake them, and I was the one that had to stay late, do another setup, submit FA every op, etc etc, for something I had no control over, and which could have been avoided entirely IMO had the guy just used a little common sense*....

elders: those who taught me tons of valuable info on machinign in general, those who some here hold in very high regard as they were doing cool/complicated shit without all the fancy cmm's, and CAM, and (modern) cnc

common sense: debatable for sure LoL... I will use this example, if you have something with a +/-.01" and you saw your guy not only checking it with a tenth mic (or whatever), but resetting multiple parts, offsetting every couple parts/tools to get it perfect, would you be mad?