ISO 8015 6

[ak762]

I would like to open new thread in continue to thread1103-196260: Tolerance analysis ISO2768
When I have read previous thread there was concerns about how to use ISO 2768-mK in correct way.
From my opinion it is sort of alternative solution to well known tolerance system x.x - x.xxx with nominal dimensions.

But ISO 8015 Technical drawings - Fundamental tolerancing principle. more cause anxiety to me.
It is stated: Each specified dimensional or geometrical requirement on a drawing shall be met independently, unless a particular relationship is specified.
As I understood from my background this standard destroy all engineering practice regarding to application limits and fits.
Designer apply fits by engineering calculation or recommendation from proved design.
If I will take into consideration fit for mating features of components I can not be sure with fit due to independently.

Are someone represent drawing under governing ISO 8015. How you overcome above mention situation?

 

[dgallup]

Isn't that the worst spec ever? Even worse than ISO 1302 on surface roughness that states something like 14% of the surface can be above the requirement!

We have our own standard called out on all our drawings stating exactly what ISO standards we use any exceptions. We did this when we were sold from one company/country to another company/country and we had to come up with a system that would not change the interpretation of all our existing drawings. This is what our standard has to say about ISO 8015:

ISO 8015: 1985 Technical drawings - Fundamental tolerancing principle

Stipulation: Clause 5.1.1 Linear tolerances: unless otherwise specified on the drawing, the envelope requirement is invoked for all drawings per ANSI/ASME Y14.5M - 1994.

Stipulation: Clause 6 Mutual dependency of size and geometry: unless otherwise specified on the drawing, the envelope requirement is invoked for all drawings per ANSI/ASME 14.5M -1994.

Stipulation: Clause 6.1 Envelope requirement: unless otherwise specified on the drawing, the envelope requirement is invoked for all drawings per ANSI/ASME 14.5M - 1994.

Stipulation: Clause 7.2 Designation: unless otherwise specified on the drawing, the envelope requirement is invoked for all drawings per ANSI/ASME 14.5M - 1994 from the American National
Standards Institute.

This is totally legit, there is a clause somewhere in ISO 8015 that says you can invoke national standards.

 

[pmarc]

Answering shortly to your last question - to overcome this situation envelope requirement (E in the circle) has to be specified in conjuction with limit values. This will basically do the same thing as Rule #1 in GD&T according to ASME Y14.5, so it will bring dependency rule into play.

 

[fcsuper]

You could add a note to your title block that invokes circle E on all dims unless otherwise specified.

Matt Lorono, CSWP
Lorono's SolidWorks Resources & SolidWorks Legion
Follow me on Twitter

 

[ak762]

I gave star to dgallup for sharing info. In two last answers was offered clever solution and obviously for me as engineer but I am in front of drawings where stated ISO 8015 without any exceptions. In one company department there is strong decision step out from Rule #1 and follow by ISO 8015.
They produce validated product. Anyway there is cooperation among other vendors which components have been installed and equipment have long lifecycle.
Unfortunately for me raised topic going outside of GD&T points. It is regarding to potential troubles what I see.
Do not wish to be fired with my review for chief because I predict at final stage we will come to what is primary governing design or manufacturing.
I would appreciate from local community pros&cons for come back to normal practice.( from design point of view )

 

[pmarc]

ak762,
As you already mentioned ISO 8015 says that "each specified dimensional or geometrical requirement on a drawing shall be met independently, unless a particular relationship is specified". Putting E in the circle next to the limit dimension or invoke envelope reuirement by general note is exactly what "unless a particular relationship is specified" means. Envelope requirement is not an exception to ISO 8015. It is allowable and recommended practice clearly legitimized by the standard (see paragraph 6.1 of 8015).

I would even say it is necessary when assembly of two features is considered. Putting only dimensional limits for these features, without invoking envelope requirement, does not assure mating at all.

 

[ak762]

pmarc,
Well input with last sentence and I have the same opinion in this matter. Personally in mechanical engineering I can not recall any component without incorporation into assembly.
But there is another interpretation and understanding of "unless a particular relationship is specified" with ISO 8015 let say (manufacturing point of view)
At component drawing designer/drafter do not specify relationship between mating features he can keep in mind it only or do not depends from input data.
"Therefore, where no relationship is specified, the geometrical tolerance applies regardless of feature size, and the two requirements are treated as being unrelated".
Manufacturing wish to interpret Principle of independency such way and delegate assure mating with secure functionality to designers only. They see cost efficiency from this concept.
I would not say it is impossible with stuck up calculations but this approach generate huge headache to engineers and anyway do not allow to achieve cost down

For example.
I am going to design steam for hydraulic valve. I chose seals. Rod diameter is recommended with h9 limit by vendor catalog.
What I will do if I go with manufacturing interpretation of ISO 8015. I have to recalculate fit for meet independency and secure functionality.
Instead h9 it can be specify h8 of finer even. Cost of part grow with precisions and at end with this approach I do not achieve any effect with cost.

Just my thought after reread several times and speculation about ISO 8015
I am wondered sometimes how wise someone who composed standards.
You can interpret information different way and it will be always right. Decision up to user without observation behind of it.
There is stated emergency exit with envelope principle even if you wish to represent drawing with Rule #1.
But may be better to clarify what is mean "unless a particular relationship is specified" or do not publish such document.

 

[pmarc]

ak762,

I understand your standpoint and I am not an advocate of ISO 8015. I am just trying to tell that being in line with ISO 8015 does not always mean you have to tighten tolerances as you described in the example. If you tie form and size of a rod together by applying envelope requirement, your functional requirements will be met by h9 fit. ISO 286-1:2010 "Basis of tolerances, deviations and fits" actually recommends envelope requirement every time when it is needed.

ISO 8015 is based on completely opposite philosophy than ASME Y14.5 and it is almost impossible to change it. ISO experts have their own arguments to make independency rule as default condition for dimensioning and tolerancing. I heard a lot of opinions on this topics and actually most of them say that Rule #1 is inpractical and increases production and inspection costs. I know it seems to be in total contradiction to ASME's approach, but believe me, they did a lot of different industrial studies to satisfy that thesis. If you take a look for instance to G.Henzold "Geometrical Dimensioning and Tolerancing for Design, Manufacturing and Inspection", chapter 17.3, you will see something like this:

http://files.engineering.com/getfil...c9-b52dc16833d2&file=Henzold_excerpt_17-3.JPG

We can agree with this or not, but there is no better evidence of the way how ISO is approaching dimensioning and tolerancing issue.

P.S. IMO ISO 8015 is very clear on "unless a particular relationship is specified" statement. In the last part of chapter 4 it is said:
"Consequently, if a particular relationship of
- size and form, or
- size and orientation, or
- size and location
is required, it shall be specified on the drawing (see clause 6)."
And clause 6 states:
"Mutual dependency of size and geometry may be called for by
- the envelope requirement (see 6.1);
- the maximum material principle (see 6.2)."
Additional explanatory pictures follow this statement.

I have seen different ISO standards on GD&T and I can tell you that 8015 is one of the clearest ones.

 

[KENAT]

To some extent this can be summarized by ASME (and historically the English system as I understood it) seeming to be more from the point of view of designers/engineers ensuring fit/function of parts at the expense of potentially causing higher reject rates.

ISO on the other hand seems to come from a manufacturing point of view of trying to allow as many parts through as possible.

Posting guidelines faq731-376

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

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

 

[fcsuper]

ASME's Y14.5 approach is formed from an ongoing series of lawsuits that helped define the rules that are now within it. In my uneducated opinion, the ISO approach will be the one that changes over time as more lawsuits come out of relying too heavily on the loose rules in the ISO system in Europe and Asia.

The ASME standard was born in fire as an industry struggled to find a place for accountability. The ISO standard was born out of European contempt for anything with the word "American" in the title.

To address another point:
I see use of the ASME standard growing in the US. The growth has exposed the lack of GD&T training in the schools.

Matt Lorono, CSWP
Lorono's SolidWorks Resources & SolidWorks Legion
Follow me on Twitter

 

[CheckerHater]

To summarize, ISO is based on Real world assumptions, and ASME based on Imaginary one.
Also, majority of manufacturers in US (even those aware of GD&T) are actually measuring their parts according to Independence principle without even knowing that.

 

[pmarc]

In my uneducated opinion, the ISO approach will be the one that changes over time as more lawsuits come out of relying too heavily on the loose rules in the ISO system in Europe and Asia.

I am not so sure about this, at least in terms of independency as a default condition. In fact it rather seems to be that ASME is leaning towards giving more freedom to usage of independency rule by introducing (I) symbol instead of note: "PERFECT FORM AT MMC NOT REQD" in Y14.5-2009.

CheckerHater,
You pointed to very important issue. It would be really interesting to investigate how often Rule#1 is verified in industry. I think results could be quite surprising.

 

[ak762]

In fact it rather seems to be that ASME is leaning towards giving more freedom to usage of independency rule by introducing (I) symbol instead of note: "PERFECT FORM AT MMC NOT REQD" in Y14.5-2009.

I proceed from old engineering practice drawing for simple part should not be complicated.
It is why I am leaning to ASME 14.5M with Rule #1 but seems this time is over and drawings will looks like illuminated X-mas tree with bubbles according any standards.
Could someone reference to Independence principle regarding to DIN

 

[CheckerHater]

I am posting from work, so I don't have much time.

But I don't want to look like a troll, so I will explain my point in short "installments"

First, take a look at the enclosed part picture.
To satisfy Envelope requirements part has to be checked with gauges, that are economically not feasible or even physically impossible to produce with today’s technology.

So, like it or not, someone will take micrometer to measure diameter (“local”, no matter in how many places you take the measurement), and then check the straightness with set of “vees” and indicator.

And ISO does reasonable, real-life thing. It sais: If you take two different measurements with two different instruments, let’s specify them on the drawing INDEPENDENTLY from each other.

(more to follow (maybe))

Link:

 

[ak762]

Thanks for sample but I have opposite opinion regarding your words.
Precision is more related to equipment and fabrication method with your sample then to inspection tools.
If it would be crucial for this component you must to use right tool for qualify part according to depicted dimension.
Up to buy advanced tool even (scan for example)

ISO 8015 allow to make a compromise. Instead of it I would revise design of part if it is not so important for function.
Combination design and manufacturing do the best product with cost efficiency. ISO 8015 slow down this process
What I like from ASME 14.5M with Rule #1 that it is pull us to progress at least.

BTW I finished my review based on this discussion. Wait feedback from chief

 

[DeanD3W]

CheckerHater,
The "no-go" gage you show is a "no-go". To use that gage would imply that there is a perfect form boundary at LMC, which for the case you show, there is not. The requirement at LMC is to measure a sufficient quantity of local sizes, with all required to be 9.9 or greater (leaving measurement uncertainty for another discussion).

ISO may be attempting to address real-life, but it is not the best approach. Unless circle "e" appears on all size tolerances there is positively inadequate control on the feature geometry. Relying on manufacturing process capability is the only savior without a perfect form boundary and mating envelope measurement. With a perfect form MMC boundary, fit can be guaranteed, or with a perfect form LMC boundary (when an L modifier is placed after the tolerance value in an associated feature control frame) the presence of material for post measurement machining or thin wall concerns can be guaranteed. Without a perfect form boundary you really have nothing... Your data is instead just a bunch of numbers that cannot predict whether a part will function or not.

Dean

http://www.d3w-engineering.com

 

[KENAT]

In fairness to ISO though they're probably presuming you'll invoke something like iso 2768 which to some extent addresses some of the 'missing tol spec info'.

However, I still have my issues with that standard including that it essentially washes it's hands of the idea of using the drawing to detail what is an acceptable part or to reject parts.

Posting guidelines faq731-376

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

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

 

[pmarc]

If we want to discuss about ISO GD&T in details we have to keep in mind a fact that even if there is no (E) modifier next to dimension value, it does not have to necessarily mean that the geometry (shape) of a feature is not controlled. In fact - according to ISO - it might be perfectly defined.

There is something like general geometrical tolerances concept which is quite common in ISO world, but little known in US. It assumes that as long as there is a reference to appropriate standard about general geometrical tolerances on a drawing there might be no need for specifying particular geometrical tolerance.

There are at the moment two standards about general tolerances:
- ISO 2768-1:1989 - "Tolerances for linear and angular dimensions without individual tolerance indications";
- ISO 2768-2:1989 - "Geometrical tolerances for features without individual tolerance indications".

As their titles suggest first part is dealing with linear and angular general tolerances and the second one with geometrical tolerances like: straightess, flatness, circularity, parallelism, perpendicularity, symmetry, circular run-out. Both documents define tolerance classes which generally depend on size of dimension. So if for instance a drawing contains a note like: "GENERAL TOLERANCES ACCORDING TO ISO 2768-mH", it will mean that tolerance class for linear/angular dims. is medium (m) and for geometrical tolerance is H. Then looking at the tables in abovementioned standards one can say what are the exact values of allowable linear and geometrical tolerances.

There is also an option to invoke envelope principle as a default condition to a drawing. Note like "GENERAL TOLERANCES ACCORDING TO ISO 2768-mH-E" will do a thing, so 'the ocean' of bubbles does not have to happen.

I have attached one example directly from ISO 2768-2:1989 that shows extreme case when very simple part does not have any geometrical tolerance applied to it and the geometry is yet fully defined.

http://files.engineering.com/getfil...-6333-4375-9ba5-0664d30cc77c&file=gen_tol.JPG

I am not saying I agree with this approach. In fact I see a lot potential problems. I just wanted to give a brighter picture of how the situation with ISO standards looks like in reality.

 

[CheckerHater]

OK, I will continue.

To AK (is 762 a caliber?): will you buy advanced/custom tool to verify every dimension for every part, or will you at some point, make a “compromise”?

To Dean: You are hair-splitting. You understand this is not real part, and you are steering away from giving the answer – exactly how will you measure envelope requirement - program million points into CMM and you get million of local measurements still.

“ISO may be attempting to address real-life, but it is not the best approach” – love it. GD&T used to be about real world (and AK, I am with you – things have to be simple), now it’s about self-preservation.

But let’s go back to cylindrical parts.
Now imagine two of them – one very short, say 30mm and another very long, 1000mm. Both still 10mm dia.

The short part will work as a dowel. As we try to hammer it into mating part we realize that variation in shape is as important as variation in size, so Envelope requirement is absolute must. In ISO world you can specify Envelope requirement on single dimension or on the entire drawing – I am beating the dead horse here.

The long part is sort of a rail. I buy pre-machined (TG&P) shaft and cut off piece of it. (To Dean – there will be some more machining and maybe heat treatment involved). When I am done I want to make sure I didn’t make it worse, so I check the specks. Original shafting was +0.00/-0.05, still good, and 1.5mm per every 1.5m straight. I cut it shorter, so I can demand 1mm straightness. Oops! Now I am in violation of ASME Y14.5. Because of the Envelope principle straightness can only be smaller than the size tolerance. Luckily, ASME now allows use of (I) sign to invoke Independence principle.

I am trying to make myself obvious: Neither (E) or (I) is good, or bad, or un-American. Both are simply tools that you apply to describe how parts are manufactured / inspected .

(more to follow (maybe))

 

[DeanD3W]

CheckerHater - I guess there will be more. No offense is intended and this forum serving a very useful purpose, considering all the disagreements regarding GD&T. You and others posting are clearly very knowledgeable. My interest is to participate when I think I can help someone else (though maybe not the poster I respond to directly) or when a discussion might lead to learning something myself (which is always).

Yes, I realized that the part you depicted was not a real part. Regardless, the no-go gage you show is not a meaningful or valid gage. I'm not sure how pointing that out has anything to do with splitting hairs. Yes, GD&T is about the real world, did I say otherwise? I like simple too. As simple as possible while still addressing what "Mother nature" presents for us to control.

Yes, there are ways to specify what is needed with both ASME and ISO. In my opinion the ASME standard has selected the better default and ASME Y14.5 provides much more in one standard which requires something like 20 standards in the ISO system. I think ASME's approach is simpler.

Per Y14.5M-1982 section 6.4.1.2 and Y14.5M-1994 section 6.4.1.2, and also in y14.5-2009, a straightness tolerance value has been allowed to be larger than the feature's size tolerance. Independency just provides a way to separate size and form in the absence of such a tolerance.

pmarc & KENAT - Thank you for the ISO 2768 information. All of these different standards... This illustrates why buying ISO 1101 plus the other standards you need costs something like $1800..? So, am I correct if I say that there is a choice between using circle E or citing 2768, if using ISO standards? Maybe the major point of this thread is that 2768 (either mH or mH-E) should always be referenced if 8015 is referenced? I see that circle E is only shown in table 2 of ISO 1101, with reference to 8015 for explanation... Can 8015 be "safely" referenced by itself, with no accompanying reference to 2768?

Dean

http://www.d3w-engineering.com