needhelp with composite box

[ronj100]

I have tried several types of composite boxes, but none seem to describe what I want, or at least they dont look right to me.
My intent is to have the .375 dia have a position tolerance of .014 from datum -a- but .003 from centerline of datum -b-.
I also want an orientation tolerance of .003 for perp to -a- and perp to -c-.
What would that

 

[ronj100]

Gee, I dunno.
I guess the no datum kinda made sense to me when I thought about it.
If there is a position callout of a feature and no datum, than the relationship is to the feature itself, not something else (like a datum).
And, of course, this can only be multilple features of the same geometry in line with each other. Golly, I wonder if you had multiple forms of any kind in alignment that such a non-datum callout would work. A composite box for form? Possibly.
Just thinking out of the book. lol.

 

[fsincox]

ronj100,
I like you already
Frank

 

[fsincox]

I my opinion those of us trying to get work done should not be subjected to punishment by the committee for years of their inability to agree and their poor examples, something like this is the perfect example. It was OK, then, some think it is not, later it is, again. They just screwed up.
Frank

 

[ronj100]

Years ago I took my GD&T classes from Lowell Foster (He is one of the authors of ASME Y14.5 1994). Only recently I returned to the field, so I am rusty on my GD&T. But I definitly remember Lowell telling us that the concepts of GD&T, if understood, had much more potential as well as symbolic representation possible than what was presented in the Standard. The Standard was just to get the ball rolling in the industry. Lowell actually encouraged us to think up additional logical ways to represent location or orientation as long as it built upon the standard. So my point is that if a GD&T callout makes sense according to the concepts that the standards are explaining, but that specific callout is not represented in the standard, is it always to be wrong?

That goes back to my question to pmarc which he has not answered. I could not find an example of option#3 after browsing through my ASME or Lowell Foster books. Anything in 1994 specifically allow or disallow combining a single segment FCF with a separate composite control frame as illustrated in option#3?

 

[fsincox]

Ron,
I was trained by another student of Foster's and was also given the impression that the standard was a work in progress and not an absolute limit definition. That was when the ISO and ASME were going to live happily ever after, in the 1994 version.
You are asking pmarc for the impossible, they did not yet see the possiblities or, at the very least, let it be known to the rest of us at that time.
I have books by committee people that discuss it and those of us that were concerned with defining functional requirements of parts saw it had to be coming.
Mark Foster seems to be of the same vein, I wonder if he is related?
Frank

 

[fsincox]

Two points,
First, to be fair a standard is by definition an attempt to keep everyone on the same page so I do understand to some extent the rigidity is required or you have no standard at all.
Second, I agree that the old examples of position without datums were in most cases just examples of bad practice, even when the early standards did it, the attempt at rectifying that in 1994 but then rectifying that new error in 2009 should be enough to see it should be disregarded, IMHO.

 

[Belanger]

I haven't been following the main discussion here, but I wonder why Frank (and others?) have a problem with a position tolerance with no datum references. The example given was Fig. 5-51 of the 1994 standard; suppose we lopped off the upper tier and just had the 0.15 position? I see no problem with that. Maybe not a wise design, but it seems legal because it just positions the holes to each other only.

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

 

[fsincox]

JP,
Sorry to confuse, I am saying it is definitely OK, IMHO.
Frank

 

[Belanger]

No problem -- I guess that's what happens when I jump in mid-stream. I'll butt out again...

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

 

[fsincox]

JP,
I am not a writer and reading back I think I can see that I may not be expressing myself completely and clearly. That is no fault of yours.
I am also actually harking back to another thread, here, that was discussing some "1994 prohibition against position without datum references" that completely through me for a loop. Thankfully, the committee has come to its senses since then! Thus my position being if it was OK, then not OK, then OK again, they most likely just made a mistake!
Frank

 

[pmarc]

@ronj100,
First of all, do not expect to find each and every case in the standard or in any other publication.

Nothing in '94 standard, as far as I can tell, allows or disallows "combining a single segment FCF with a separate composite control frame". However, in my post from 4 Sep 13 16:01 I gave you one example, where this combination would be the only way to grasp functional intent. In my opinion this solution would not violate any of GD&T rules defined by the standard.

Nothing in '94 standard, as far as I can tell, allows or disallows using single segment FCF without datum feature references too. However, since there is no example in the standard showing its applicability, yet there are examples showing that position without datum feature references can be used in composite callouts, I prefer option #3. Like I said, there are many people, sometimes GD&T authorities even, claiming that when something is not explicitly shown in the standard, it is illegal. I am not amongst them, thus I also said you can choose option #1 or option #3 depending on your preferences.

As for stand-alone positional callouts without datum references applied to patterns of features (that is where there are no other geometric tolerances controlling location of the pattern), I am all for it as long as the pattern serves as primary datum feature. (Again, this concept is not shown in '94 standard, but is explained in '09 edition). In any other case, this would be incomplete drawing specification. Location of such pattern through directly toleranced linear dimension(s) is very far from naming it "clear & unambiguous".

 

[fsincox]

I believe the practice in pmarcs last paragraph is all the committee is trying to get away from. A practice you will see alot on old drawings and, as I pointed out, in the older standards themselves.
Frank

 

[ronj100]

Thanks pmarc,
I really appreciate the clarifications and explanations. They help much.