ROUNDING OR NOT, DIMENSION, TOLERANCE 2

[dho]

Lest say a dimension, .040+/-.001. The part is .0411. Good or Bad ?
Please enter your VOTE.
We have a big fight here !!!

 

[KENAT]

What drawing standard do you work to?

At least if ASME Y14.5M-1994 [section 2.4] (and as far as I know other versions of that std) all limits are absolute. "Dimensional limits, regardless of the number of decimal places, are used as if they were continued with zeros."

So .040+-.001 means the upper limit would be .0410000000000000000000000000000000000000000000000...0 and there is no need for any voting.

Of course, I suppose potentially someone could argue your measurement error could come into play but generally that is taken to reduce the available tolerance not increase it.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[fsincox]

Certainly, bad by the book!
Frank

 

[dho]

thanks a lot.
now i have something to go back to argue with our QA manager who told me that he have 40 yrs experience.

 

[CheckerHater]

No need to ask "What drawing standard do you work to?"

It's exactly the same, ASME or ISO. Infinite number of zeros follows. ;-)

 

[KENAT]

CheckerHater, those aren't the only 2 options. The answer could have been some internal company standard or similar that set's its own rules.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[CheckerHater]

Agree

 

[fsincox]

Ken,
If it is measurement error, should it not be remeasured and the correct data entered it its place?
Frank

 

[jbeckhou]

Good and Bad

Petrotrim Services

http://www.petrotrim.com

 

[KENAT]

fsincox, inpspection error budget or whatever you preferred term is if you want to be picky. I was talking about the "10%" of tolerance that inspection typically uses.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[looslib]

The part is bad per the print.
The part still being useable depends on the function in the assembly.

The QA manager's 40 years experience and $4.50 will get him a Starbuck's, but won't make a bad part good.



"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[dgallup]

That's a bad part in any Quality System system I've ever worked to, ISO, ANSI, proprietary home grown. Will it work? That's a completely different story and we don't have enough information.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[TheTick]

Put the part in the freezer for an hour then measure.

 

[thepete]

Although I agree that it is bad, here's a different way to look at it. If the spec has 3 decimal, why are you inspecting to 4? And if you moved to a less accurate 3 decimal measurement, wouldn't you effectively be rounding down?

 

[ajack1]

Clearly the QA manager is wrong, however if he has been passing parts like this for the last 40 years and it has had no downside like parts failing or not working correctly then the tolerancing is wrong as well, or at least more tightly tied up than it needs to be.

So it is perfectly possible to design a part wrong, manufacture it incorrectly and inspect it incorrectly but have no negative downside over a 40 year period, all of which goes to prove it is better to be lucky than good.

 

[dgallup]

thepete - The inspection system is supposed to be 10 times more accurate than the tolerance.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[pmarc]

Nothing is perfect in reality.

A part can't be made perfect, and a measurement is never perfect - certain amount of measurement uncertainty will be always coming into play. And this uncertainty may make a huge difference, especially in cases like this one.

So from purely theoretical point of view ".0411" means the feature/part does not meet drawing specification. However, if the measurement uncertainty has been considered, the verdict may not be that obvious.

 

[CheckerHater]

The part is bad; there is no uncertainty to that.

So it is perfectly possible to design a part wrong, manufacture it incorrectly and inspect it incorrectly but have no negative downside over a 40 year period

The case on hands is what ISO calls Correlation Ambiguity, meaning, in plain words, that part specification does not represent functional requirements properly.

 

[dho]

thanks to all.
our QA man was wrong on this. he was saying we could accept up to .04149999....9.
everyone in my eng dept do not agree. but can not remember where the PRINT is.
now, we have ANSI Y14.5 section 2.4.
we understand the accuracy requirements for the measuring instruments and the whatever the uncertainty of it.
thanks again.

 

[ctopher]

It's bad per dimension tolerance.
Good to whoever doesn't care, or know what it means.


Chris
SolidWorks 13
ctopher's home
SolidWorks Legion