ISO GD&T vs. ASME GD&T 8

[pmarc]

What do you think about discussing more in details about differences between ISO and ASME GD&T? What are the most important discrepancies between the two in your opinion?

 

[KENAT]

While others may be able to articulate it more fluently, one trend seems to be that ASME is about defining the functional design requirements and what will be accepted, while ISO seems to lean more toward minimizing rejections/maximizing 'good parts' with less regard to function.

Posting guidelines faq731-376

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

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

 

[pmarc]

Kenat,
Here is a content of paragraph 4.1 from ISO 1101:2004 - one of the most important standards from the group of ISO standards:

"Geometrical tolerances shall be specified in accordance with functional requirements. Manufacturing and inspection requirements can also influence geometrical tolerancing. NOTE Indicating geometrical tolerances on a drawing does not necessarily imply the use of any particular method of production, measurement or gauging."

Doesn't this look similar to Y14.5's fundamental rules 1.4(d) & (e)?

Isn't a general purpose of GD&T (regardless whether it is ISO or ASME) to minimize rejections / maximize "good parts"?

 

[KENAT]

That's what they say there, elsewhere they say other things.

Posting guidelines faq731-376

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

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

 

[pmarc]

'Elsewhere' means where?

 

[Belanger]

How 'bout the fact that the ASME standard is one volume with a cost of US$170, but for ISO you have to buy scads of separate standards where the total cost gets into several hundred dollars!

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[ctopher]

thread1103-228088

Chris
SolidWorks 10 SP5.0
ctopher's home
SolidWorks Legion

 

[KENAT]

Well, you could checkout some of the infamous last paragraphs of ISO2768 (I think part 1 but it may be in part 2 as well, just thinking about that damn standard makes me want to punch whoever wrote it so I'm not going to go look it up).

Posting guidelines faq731-376

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

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

 

[CheckerHater]

I recently stumbled upon the following quote:

"All features on component parts always have a size and a geometrical shape. For the deviation of size and for the deviation of the geometrical characteristic (form, orientation and location) the function of the part requires limitations which, when exceeded, impair this function.
The tolerancing on the drawing should be complete to ensure that the elements of size and geometry of all features are controlled, I.e. nothing shall be implied or left to judgment in the workshop or in the inspection department."

Could somebody identify if it comes from ISO or ANSI/ASME standard, or somewhere else?

 

[TED7]

Kenat,

Are you refering to ISO2768-Part 1 section 6 (and clause A4) along with ISO2768-Part 2 section 7 (and A4 again)

"Workpieces may be whatever shape the manufacturer decides, the designer has no say in the matter unless it's totally fubar"

Has to be my favourite part of any standard

Apologies if you punch anyone as a result of me looking it up for you.

Designer of machine tools - user of modified screws

 

[TED7]

Checker, as it's on my desk, your quote is also ISO2768, its the introduction to parts 1 and 2.

Designer of machine tools - user of modified screws

 

[CheckerHater]

Finally someone in the possession of actual book!

Ninja, could you back me on the fact that 2768 is very limited in scope, as it only applies to dimensions without tolerances usually typical for "non-critical" features?

And in fact, it doesn't apply to "invisible dimensions", such as hidden inside of 3D-model?
And also it doesn't apply to reference dimensions and basic dimensions?
And also it doesn’t cover cylindricity, profile of any line, profile of any surface, angularity, coaxiality, positional tolerances and total run-out?

And it still your right (and responsibility) as a design engineer to:
Explicitly indicate smaller tolerances than 2768 on the drawing when required for proper function?.
Explicitly indicate larger tolerances than 2768 on the drawing when permissible for economical manufacturing?
Explicitly indicate that your favorite (sorry, favourite) section 6 (and clause A4) do not apply?

 

[MechNorth]

Within ISO, Rule #1 does not apply (unless specifically invoked) therefore size does NOT control form of a FOS. In other words, within the ISO system you typically control form separately from size. ISO 2768 deals with the size of a feature, and by extension of the preceding, not its form. I've heard arguments to the contrary, but nobody's backed it up from the ISO 2768 standard. Now, to be specific, the general tolerance classes in 2768 apply to non-toleranced dimensions which are NOT reference and which are NOT basic. Reference dimensions don't have any meaning as far as part definition is concerned. Features established with Basic dimensions typically are controlled with feature control frames. A notable exception for basic dimensions occurs when they are used to locate and/or size a datum area target in which case a basic dimension has gage-maker tolerances applied.

Another difference between ISO and ASME is that ASME tolerances are nested, or inclusive. As you work down a stack of controls on a feature, each subsequent control must be a refinement of the preceding control (i.e. a smaller tolerance zone). With ISO, each control is considered separately and therefore you could for example have a parallelism of 0.1 with a flatness of 0.5. In other words, in ISO the tolerances are cumulative. As an extension, within ASME the surface finish/roughness is considered a refinement of the surface form; within ISO, the surface roughness is cumulative, so it can actually exceed the flatness.

Overall, ISO seems to be moving to a metrology-based standard. The definitions of the various controls indicate means of checking the control rather than focusing on the functionality of the control. ASME has always been functionality based and with 2009 moved even further from including any instructions on verification.

The differences between the standards is a big topic.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[Belanger]

JIm, I've seen that mentioned before (about ISO thinking of tolerances separately, rather than refinements). But I'm having trouble picturing your example: if I verify a surface parallelism of 0.1, how the heck could it's flatness error be more than 0.1? The logic just doesn't connect. (Unless ISO's parallelism is something like tangent plane or each element.)

Also, I don't think ASME sees surface roughness as a refinement of form. You could have a crowned surface that fails flatness, yet its surface roughness could be as smooth as glass.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[pmarc]

Jim,
It is interesting what you are saying about relationship between values for parallelism vs. flatness tolerance (0.1 vs. 0.5). I know you used this as an example only, however I am really tempted to not agree with you.

Paragraph 5.2.2 of ISO 2768-2:1989 "General Tolerances for Parallelism" states that: "The general tolerance on parallelism is equal to the numerical value of the size tolerance or the flatness/straightness tolerance, whichever is the greater."

Doesn't this mean that parallelism value must be at least equal to flatness value?

 

[pmarc]

J-P,

Also, I don't think ASME sees surface roughness as a refinement of form. You could have a crowned surface that fails flatness, yet its surface roughness could be as smooth as glass.

This is because actual surface roughness is a refinement of actual flatness error. But the opposite situation can't happen, and this is, I believe, what Jim meant.

 

[Belanger]

No, the opposite situation is what I meant. In fact, both cases are possible: A surface can pass surface finish and fail flatness.
Also, a surface can fail surface finish and yet pass flatness.

I guess the confusion is because the parameters for surface finish are localized (the cutoff width), but flatness spans the entire surface all at once.

So in this regard, both ASME and ISO see flatness and surface finish as separate requirements, which was my intent in replying to Jim.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[CheckerHater]

Jim,

Allow me to disagree.

"Within ISO, Rule #1 does not apply (unless specifically invoked)"
All you need to invoke Rule 1 in 2768 is to add (E) symbol to the call-out, just once on the entire drawing sheet. ISO is well aware of Envelope principle and puts it to good use.

"ISO 2768 deals with the size of a feature, not its form. I've heard arguments to the contrary, but nobody's backed it up from the ISO 2768 standard"
ISO 2768 covers Flatness, Straightness add Circularity which ARE form controls. Nobody will "back" it for you, open the book and see for yourself.

"With ISO, each control is considered separately and therefore you could for example have a parallelism of 0.1 with a flatness of 0.5"
Entirely not true. In ISO location and orientation tolerances also control form characteristics. In your case flatness may be indirectly controlled by the Envelope requirement, Perpendicularity, Parallelism, Angularity, surface Profile, Position, Total run-out, and general tolerances, like ISO 2768.

"As an extension, within ASME the surface finish/roughness is considered a refinement of the surface form; within ISO, the surface roughness is cumulative, so it can actually exceed the flatness"
Wrong again, flatness specification could limit surface waviness or roughness, it's just no one cares because form tolerances are much larger that surface texture specs.

 

[CheckerHater]

I should refresh the page before posting :-(

 

[TED7]

ISO2769 should only be used to save the designer the hassle of tolerancing every feature on the drawing. That is its only purpose. It's a more useful version of "all dimensions +/- 0,25mm" no more

Doesnt mention dimensions in a 3d model (the standard is from 1987, its likely this concept didnt exist at the time).
Doesnt apply to reference (auxiliary in the standard) or basic dims.
Doesnt cover cylindricity but implies cylinders are controlled by circularity and paralelism, which the standard does cover. Profile of a line and surface are covered lightly by straightness, flatness and paralelism. Angularity no but it covers tolerances on angular dims. Coaxiality is not covered but coaxiality may be limited by circular runout which is covered. Positional tolerance and total runout are not included.

The last three are correct, what use is a designer who doesnt control tolerances which improve the design. The last bit of the standard is right when interprited correctly and common sense applied, it needs re-phrasing badly.

Designer of machine tools - user of modified screws