Single Datum Axis from 2 Coaxial Datum Features

[dtmbiz]

Is it possible to identify 2 coaxial features of size (cylinders) @MMC for both to, establish a single Datum axis ? (e.g. A-B)

 

[pmarc]

Yes, it is possible. See figs. 4-24 and 7-59 in Y14.5-2009.

This approach can be also used for coaxial cylinders which are not of the same size. Assign one of them A, the other one B, and the reference to the common datum axis will then be A(M)-B(M).

 

[dtmbiz]

pmarc

I am specifically asking about the "A-B" concept. Fig 4-25; where Datum feature simulator ("simulator" spelled wrong in 2009 figure)
I probably wasn't clear enough.

I do not believe MMC is valid in this scenario.

Thank you pmarc

 

[powerhound]

Yes, it is valid only if you are controlling the considered feature with position or profile. If you are controlling the considered feature using runout then you cannot use the MMB modifier.

EDIT: It is definitely not valid in figure 4-25 since the considered feature is controlled by runout.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[dtmbiz]

powerhound


For A@MMB-B@MMB (position control) there would be a gage built with hole sizes equivalent to each features MMC ?

If the feature is smaller than MMC then there would be datum displacement (wobble within the gage) ?

And from that, how is a "single axis" defined? The axis is from the features' coaxial holes @MMB ?

 

[axym]

dtmbiz,

I agree with pmarc and powerhound. Referencing two coaxial datum features at MMB is possible. The A(M)-B(M) case is not illustrated in a figure, but I don't see a problem with it. This would not be possible for Fig. 4-25, but only because of the special rule for runout tolerances that the datum features must be referenced RMB only.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[dtmbiz]

Thank you to all responders

 

[powerhound]

Sounds like you got the answer you needed but since you asked me a question right there at the end, I'll answer it as best I can.

And from that, how is a "single axis" defined? The axis is from the features' coaxial holes @MMB ?

The datum axis would be the axis of the aligned datum simulators.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[axym]

dtmbiz,

As powerhound said, the datum axis is the axis of the two datum simulators (in other words, the datum axis is defined in the gage). Having said that, the part can wobble (rotate and translate) within the gage. This is one of the tricky things about datum features referenced at MMB - from the part's point of view, there are many different possible datum axes (one for each different "wobble configuration" of the part in the gage). There isn't one unique datum axis. Imagine the part as stationary, and the gage (with its datum axis) wobbling relative to it within the clearance. That's the situation.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[greenimi]

This approach can be also used for coaxial cylinders which are not of the same size. Assign one of them A, the other one B, and the reference to the common datum axis will then be A(M)-B(M).

pmarc, Evan, John,

Can we even assign one of them as A and being controlled/positioned/ located to A(M)-B(M), the other one B and being controlled/positioned/located to A(M) -B(M) and the reference to the common datum axis will then be A(M)-B(M). Am I correct or some caveats are needed?

 

[dtmbiz]

Axym
You stated exactly my concern. With "many axis" possible, it doesn't seem to fit my definition of "a single axis".
Yes each axis is single, however there isn't a "single repeatable axis" as I see it.

pmarc
There are many concerns I have with this dimension scheme. For what ever reason, I do not like the looks of it.
One thing I definitely do not like its that both A & B are a length of .038 mean value.

I do not have access to mating components.

see attached

 

[pmarc]

dtmbiz,

In my first post I did not suggest to control datum features A and B with position to A(M)-B(M). Position with no datum feature references would be fine (as shown in the standard). I meant that the toleranced feature should be controlled to A(M)-B(M).

If your concern is that datum features A and B are short, then the question should be asked if from design standpoint they are good candidates for datum features at all. To me, even though they are short indeed, their separation looks to be sufficient to make them feasible primary datum features.

 

[greenimi]

For me the shown drawing looks good (correct method to express a very common true design requirements within the definitions and rules on the Y14.5) except a small "mistake": NO diametic symbol needed for the perpendicularity callout on the datum feature C.

 

[dtmbiz]

pmarc
understand and agree

greenimi
FYI
Agreed it is technically correct.
I do not like the MMC due to thin wall tube with tolerance stack
I am not sure how to factor in the datum displacement in this scenario, considering that
there is more than likely "wobble" built into it. This wobble can potentially and probably
will be different at both ends of the part.

BTW
This is not my dimensioning scheme

 

[AndrewTT]

I do not agree that the drawing is technically correct. The chapter on Runout is clear that "Runout tolerance may be applied to a datum feature(s) and related to the datum axis derived from that datum feature(s)" (section 9.5.5 - 2009). I can't find anything in the chapter on Position that allows this same concept. Said another way, The ∅.3125 features cannot reference themselves in their FCFs.

 

[mkcski]

I agree with AndrewTT - runout is different than position in this area. Although I understand the concepts with datums with runout; from a practical mfg and QA standpoint, good luck with establishing the datum axis from the entire surfaces and then checking the FIM runout.

Certified Sr. GD&T Professional

 

[greenimi]

AndrewTT and mkcski,
You might not agree that the shown drawing is technically correct (except the Ø symbol), but that does not mean is not.

Number#1:
pmarc, Evan and John are the best around here. Read some other threads and convince yourself.
To successful disagree with pmarc or Evan (to name just a few) is EXTREMELY hard and painful. You “almost” have no chance to win.

Number#2:
If the concept is not shown in the standard that does not mean is not correct or legal.

Number#3:
Quote:” The ∅.3125 features cannot reference themselves in their FCFs.”

A classic common misunderstanding in the industry = is that both datum features A and B are referencing a location back to themselves. However, that is not the case. The datum reference frame is simply specifying that the “Datum Axis” a single axis, is to be established using both datum features A and B together at the same time.

 

[greenimi]

I did not say that the shown concept is FUNCTIONAL,--I have no clue how is part works-- (however, as shown, is most likely reflecting the design intent, if the two “rather small” surfaces works together to align/orient the part in assembly), but for sure is legal.

“The figure illustrate the most practical functional requirement where both datum features A and B will be used to together to hold the part in an assembly. The two datum features are still allowed some misalignment in both location and orientation. The amount of misalignment permitted is often specified by bearing manufacturers and other design guidelines.
Because of such said above common misunderstanding, the intended coaxial control of the datum features to themselves as well as the considered features controlled with total runout/ position is often specified relying on a FUNDAMENTAL APPLICATION of the GD&T language, but lacks the advanced understanding of what the permissible variation may actually be.”

Cite / quote from Applied Geometrics Advanced GD&T book, author Mark Foster.

 

[mkcski]

greenimi:

Yes, disagreeing with the "masters" is not recommended. But to me, part of the learning process is to reflect your understanding back "at" the master and let him "correct" any misunderstanding so as to obtain more clarity. What really matters is the tact used to respond to the master. Challenging or arrogant will get you "slapped" (and detention hahaha).

A classic common misunderstanding in the industry = is that both datum features A and B are referencing a location back to themselves. However, that is not the case. The datum reference frame is simply specifying that the “Datum Axis” a single axis, is to be established using both datum features A and B together at the same time.

I have a challenge - find an example in Section 7 (or any other material) where a feature is positioned to itself (like runout allows)

I mentioned "practical" in my earlier post. So, considering fig 4-25 in the lower "means this" frame - how do you inspect the runout on the surface of the datum feature with a dial-indicator (or other device) when the datum simulators that establish the datum axis are covering it up?

Certified Sr. GD&T Professional

 

[AndrewTT]

Greenimmi - Only one of of the people that you mentioned commented after the drawing was added to this thread (pmarc). I'm not sure that one can take the support they (pmarc, powerhound, axym) gave for the OP's original question and assume that it carries over into support for the drawing that was shown later in the thread.

In my first post I did not suggest to control datum features A and B with position to A(M)-B(M). Position with no datum feature references would be fine (as shown in the standard). I meant that the toleranced feature should be controlled to A(M)-B(M).

I don't want to put words into anyone's mouth, but I read that to agree with what I stated. Hopefully pmarc can come back and clarify his position.

I am open to the possibility that I am dead wrong. But like mkcski pointed out, there does not seem to be any direct support for this in the standard.