Parallelism on per unit basis

[powerhound]

Hey folks,

Another question has come up. I found another thread on this subject but didn't find a definitive answer in it. I don't find direct support for parallelism on a per unit basis. This seems to be a legitimate extension of principle but I'd like to get others opinions or maybe point me to the place in the 2009 standard where it exists. In reality we'd like to add a tangent plane modifier to it but that shouldn't make a difference in the answer.

Thanks,

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2010
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[Belanger]

The only difficulty I can perceive in parallelism on a per-unit basis is of a practical nature. Think about it: if there are any bumps whatsoever on a surface, as the local plate which simulates the tangent plane slides around, it may rock to many different angles, and some could be quite severe if the local area is small enough.

That said, I don't see a problem with it from a theoretical point of view, so I agree that it's a valid extension of principles. But do consider the practical issues that may arise.

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

http://www.gdtseminars.com

 

[CheckerHater]

Shortly, why applying parallelism per unit is troublesome:

1. It’s illegal both in ASME and ISO to begin with.
2. Flatness and parallelism are already misunderstood and often used interchangeably (not always correctly).
3. Hard gaging is simpler for “datumless” controls where feature is measured against itself.
4. CMM programs do not always follow standards as they are, not to mention extensions.
5. After you overcome all of the above the result is still useless as you must specify “global” parallelism anyway to avoid accumulation, and flatness will provide better control to prevent abrupt surface changes (BTW, it will not prevent, but merely reduce them)

We are not talking about tangent per unit here.

 

[pmarc]

Thanks for reply, CH. Here are my comments:

1. Not true. Nowhere in Y14.5 it is said that it is illegal. Please point me to the statement saying so.
ISO 1101, 1983 edition, actually showed parallelism per unit callout as an example of callout containing a refinement in lower segment of FCF. In next editions of the standard (2004, 2012) it was replaced by straightness callout, however in my opinion that does not mean it is forbidden. Actually with the lack of a direct ISO standard on parallelism, I do not think we can judge whether it is still legal or not.

2. That proves nothing. Even assuming that both per unit characteristics are clearly and precisely described in the standard, there is always a place for misunderstanding and a risk that someone will apply it interchangeably. Similar to ASME's position and symmetry. Both are clearly defined in Y14.5, but still lots of folks use symmetry instead of position without awareness about differences between the two.

3 & 4. Again proves nothing. You focus on inspection method, not on the pure meaning of the characteristic. The fact that certain method is not suitable for checking a characteristic does not mean that there is no other method able to verify it. Actually I am curious how you imagine verifying even a typical parallelism callout by a hard gage.

5. Am I missing something or there was at least one vote here stating that unit basis parallelism worked very well in the application field? And I do not think we were ever talking about getting rid of "global" parallelism. Per unit parallelism should always be applied as a refinement of "global" callout.

 

[MechNorth]

Pmarc, another poster used it, and a friend of mine uses it regularly. Not sure that a "first level" parallelism control is not already established by a size tolerance or profile tolerance locating the surface.

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

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[Belanger]

CH -- none of those points really holds.

1) Where does ASME Y14.5 say that it is illegal?
2) So what if people misunderstand the two? Let's get them to learn GD&T, rather than dumb down a print to suit the folks that aren't getting it.
3) Not sure why you think this. Parallelism is probably easier to measure than flatness, because you can simply set the part on an inspection plate and then run a height gage across the top. To check flatness involves carefully leveling the part first, or else drilling a hole in your inspection plate (!) to mount a dial indicator through the bottom. Also recall that circularity is also "datumless," yet it is one of the more difficult controls to measure.
4) So what? Let's get the CMM people to learn the proper rules, or let them go out of business.
5) It was assumed that the per-unit parallelism was already refining a parallelism tolerance. Plus, the size dimension would already have been controlling parallelism.

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

http://www.gdtseminars.com

 

[powerhound]

My comments on your list, CH,

1. Not true, plain and simple.
2. This is not a good reason to change something legal into something haphazard.
3. How would you hard gage a parallelism callout anyway?
4. Another bad reason to change anything.
5. Per unit specification allows for using a maximum value overall.

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[CheckerHater]

Guys,

Per unit parallelism DOES NOT EXIST either in ISO or ASME. Prove otherwise.

Checkers...

 

[powerhound]

CH,

Do you even know what an extension of principle is? If so, have you ever used it? If you have, what was the application? I'm interested to know how you justified using it since your exact application of the principle would not have been shown in the standard.

http://www.tec-ease.com/gdt-tips-view.php?q=128

This tip shows how concepts shown in the standard can be used in ways not specifically in it. See "EACH ELEMENT" and the application of the between symbol for starters.

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[Belanger]

CH -- a quick question:
Are you ready to defend the notion that anything not stated in the Y14.5 standard is automatically illegal and thus absolutely forbidden?

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

http://www.gdtseminars.com

 

[MechNorth]

Here's something to consider if people don't believe in extension of principles; without them, there would be no evolution of the Y14.5 standard...period. Datum translation modifiers were added to reflect ACTUAL applications; the underlying principles were already there, but there was nothing specifically documented for this extension, despite being widely used in practice. Movable datum targets ... same thing. Profile refinement on a per-unit basis ... extension of form per unit area (e.g. flatness) to a non-datum-referenced composite profile control; enough people did it and it was not illegal by the standard, therefore the standard evolved. Extension of principles is valid and often necessary in anything but the simplest of parts. In the absence of a conflict with some part of the Y14.5 standard, an extension of principles, duly supported by established precedence in the standard, will be valid. That doesn't necessarily mean it is practical, or that it is the best way of doing something; that is to be determined by an effective review.

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

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[CheckerHater]

Powerhound,

Principle: what you think you know for sure.
Extension of principle: what you make up when you don’t know the answer.

If you have better definition please educate me: is it somewhere in ASME standard?

Belanger,

There is no need to prove negative – it’s logical fallacy.

MechNorth,

“Per unit” by itself is extension of principle – it is not supported anywhere in ASME Y14.5.1M-1994. I am waiting for new edition.

What I mean (and I stick to it) that there is absolutely nothing in ASME Y14.5-2009 to support your point of view. Please take a look:

Section 5 Tolerances of Form:

Para. 5.4.1.3: “Straightness may be applied on a unit basis as a means of limiting an abrupt surface variation within a relatively short length of the feature.”

Para. 5.4.2.2: “Flatness may be applied on a unit basis as a means of limiting an abrupt surface variation within a relatively small area of the feature.”

In both cases method and its limitations are clearly described in writing, with appropriate illustrations. No other geometrical control has “unit basis” attached to it, not even any other Form control.

Section 6 Tolerances of Orientation:

Para. 6.5: “Where it is desired to control a tangent plane established by the contacting points of a surface, the tangent plane symbol is added in the feature control frame after the stated tolerance.”

And later (in small print):
“NOTE: The tangent plane symbol is illustrated with orientation tolerances; however, it may also have applications using other geometric characteristic symbols where the feature is related to a datum(s)”

This is what you call “an extension of principle”: tangent requirement may be used with other controls, even if it is not described in detail or illustrated in the standard.

So far standard is well thought-off and balanced: Unit basis is applied to controls without datums (and not even all of them). Tangency is applied to controls “where the feature is related to a datum”.
Nowhere in the standard there is a mention of mixing two together.

Now short “Executive summary” for those not reading long posts:
1. Straightness MAY be applied on a unit basis
2. Flatness MAY be applied on a unit basis
3. Tangency MAY be applied to parallelism and other orientation tolerances.
4. Tangency MAY be extended to other controls where the feature is related to a datum(s)
5. Anything else is figment of your imagination.

I have no much interest in arguing with people in denial.

Now I am very curious about the eye-opening experience, the epiphany that made pmarc change his opinion to exactly opposite.

So far his logic is a bit flawed: “ISO 1101, 1983 edition, actually showed parallelism per unit callout as an example of callout containing a refinement in lower segment of FCF. In next editions of the standard (2004, 2012) it was replaced by straightness callout, however in my opinion that does not mean it is forbidden”

Let say, if I get rid of my Ford and buy a Chevrolet; it doesn’t mean anything; I am still a Ford owner. Convincing.

So pmarc, your current opinion is exact opposite of what you expressed in this forum:

http://www.eng-tips.com/viewthread.cfm?qid=269210

What happened? Apparently you‘ve learned something I have no idea about. Please share.

 

[pmarc]

CH,
I have said to myself that I will no longer take part of this discussion, however in this case I have to reply.
Yes, I changed my opinion about legality of parallelism per unit area through all these years, and now I am saying it is perfectly applicable as an extension of principles. What is wrong with that? At least I was willing to understand the whole concept and was open for that change.

I will repeat again: throughout the thread you have not presented a single logical argument that would prove the parallelism per unit or parallelism per unit modified by T could not be used in real life application. By "logical" I mean, a sketch or a description that would defend your standpoint. Instead of that you are using some quasi-logic arguments in order to prove that we are mistaken and not that you are right.

 

[CheckerHater]

Pmarc,
No need to be offended. You said that you changed your mind about the issue over the years.
So, what kind of evidence did you find over those years that convinced you? May be it will convince me?
So far it was you who didn’t offer any proof. At least I uploaded some sketches.
Here is another sketch, just for you.

 

[CheckerHater]

Belanger, and others,

You can prove, as in provide the evidence, that something happened sometime someplace (kind of like they do it in court).
Can you prove that nothing happened?
This is what I mean by “negative” and this is why you are assumed innocent until proven otherwise.

So it is not my job to prove that anything not in the standard is automatically forbidden.
It is your job to prove that something not in the standard is automatically permitted.

In my post I mentioned small-print note that allows extension of tangency principle to tolerances that reference datum(s). I can cite that note to defend my decision to use tangent profile or tangent position (hey, it’s extension of principle).

What do you have to show?

 

[Belanger]

All I asked is if you are willing to defend the stance that something not explicitly stated in Y14.5 must automatically be illegal. Yes or no?

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

http://www.gdtseminars.com

 

[pmarc]

Sorry CH, but this and the other sketch proved really nothing. You really even don't want to try to see the clue of the discussion, do you?
And I do not need to prove anything - the other guys' words (Dean's, Jim's and his friend's) speak for themselves.

As for what convinced me to the concept - I simply started to look at Y14.5 standard from wider perspective. It is simply impossible to gather each and every tolerancing case within a ~200 pages book. I learned that if something is not in the standard, that does not automatically mean it is illegal. If a concept is not in obvious conflict with the letter of the standard, I see no reasons to reject it just like that. Parallelism per unit area or parallelism per unit area modified by T are just the examples.

PS.: I am pretty sure you know that too. I am just surprised that you act so against it.

 

[CheckerHater]

To Belanger:

Yes.

I will present my argument later - after all I am supposed to be working right now.

 

[powerhound]

CH,

I'm really trying to see your point but your drawings are leaving me in the dark. It makes me think that we are arguing two different things. I have no idea what your most recent drawing is trying to say. It's almost like you're trying to teach us what tangent is. I'm sure not confused about that. I know what tangent is and I know what the Y14.5 definition of a tangent plane is; so, why are you providing drawings that show things that are tangent and things that are not tangent? I really don't think anyone is confused about any of it. From what I'm gathering, the sole disagreement is whether or not Parallelism per unit, with a tangent plane modifier, is okay to do.

I noticed that you made no comment on the tip. I guess Don is either confused or just making things up too. You should give him a call and correct him and make drawings of tangent and non-tangent lines so he can realize the error of his ways. You claim that we're just making this stuff up but your asinine definition of extensions of principle is the most blatantly made up thing on this thread so far. You claim we're all in denial...whatever. It reminds me of the famous Joe Walsh lyrics "It's tough to handle this fortune and fame, everybody's so different, I haven't changed."

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[MechNorth]

CH, what you found in the fine print is part of the established principle, not an undocumented extension. Once documented, an extension becomes a principle. When Einstein started working on his Special Theory of Relativity, it was far out there vs traditional thinking, but still rooted in the principles of physics and mathematics; it was an extension of established principles. Once reviewed and published, it was an accepted principle upon which new extensions could be based. Nothing different in this situation.

The concept of "extension of principles" is accepted by the writers of the ASME standards; it is why the section on Profile (for example) has grown so significantly in '09. It wasn't that people suddenly thought "hey, let's do this just because"; they brought the new material to the table based on what was already happening on industry, based on how significant companies and advanced thinkers in the field of GD&T had extended the principles. Look at composite profile tolerances for a pattern of features; it extended from composite profile controls on individual features, melded with the use of composite position controls for patterns of features. It was an extension of established principles, now an accepted principle.

Unfortunately, the sad reality is that if someone does not believe in the premise that extension of principles is the act of filling the great unknown, then there is no argument that will change the mind. I journey into that realm whenever I have to, and so far that's been more than a half-dozen times. Perhaps you have been fortunate enough to not need to push the boundaries, or perhaps you were guided to trade off functionality for simplicity. For whatever reason, you are evidently not willing to accept the concept of extending principles into new realms, at least for GD&T.

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

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[MechNorth]

Guys,
Attack the argument, not the person. Keep this on a civil basis. Certain words convey tone and insult, and are inappropriate in a technical discussion.

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

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com