×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Are you an
Engineering professional?
Join Eng-Tips Forums!
• Talk With Other Members
• Be Notified Of Responses
• Keyword Search
Favorite Forums
• Automated Signatures
• Best Of All, It's Free!

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

#### Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

# Feature Of Size definition

## Feature Of Size definition

(OP)
My first question is - according to ASME Y14.5 2009, how would you classify the cylindrical interrupted surface of diameter 55 and the width 52 in the following sketch?
Are they:
- Regular features of size (with interruptions)?
- Irregular features of size type A? (Or maybe even B?)

My second question is for those who have access to the 2018 standard:
What is the change that was introduced to the concept of feature of size?
I read that there was a change in the concept in the announcement at the ASME website which pmarc linked to in the thread about the new standard.

### RE: Feature Of Size definition

I realize I might be wrong here, but since I am the first commenting on this subject I have to live with that fact.
I would say that Ø55 feature is an irregular FOS because you can drive an UAME from it.
Not sure if anything changed conceptually about FOS in 2018.

### RE: Feature Of Size definition

I think it depends on the geometric control you're planning on using. If it's a profile tolerance to control the 'feature' defined by both the 52 and 55 dimensions then I would think it's an Irregular feature of size, type B. I say this because the angled and planar elements of the feature are not contained within the boundary of a an actual mating envelope that is a sphere, cylinder, or pair of parallel planes.

There's not enough information to say for sure. You'd need to know exactly what the extents of the feature are.

I'm not a vegetarian because I dislike meat... I'm a vegetarian because I HATE PLANTS!!

### RE: Feature Of Size definition

I agree with greenimi - you can obtain a repeatable UAME from the interrupted surface. I am waiting on my copy of 2018 to arrive. ASME says delivery Friday 3/8.

Certified Sr. GD&T Professional

### RE: Feature Of Size definition

Sem,

Your 55 diameter dimension is an interesting case. While it is not a complete 360deg cylinder, it does not look like your typical case of say a 180 deg arc which is comprised of BOTH opposed and unopposed elements - it seems from the way it is shown that it is comprised completely of only opposed elements. I could see an argument in that case that it is a regular FOS - the less controversial claim of course being that its an irregular FOS type A for the reasons mentioned by others.

In regards to the 52 width, I'm not sure that it is a FOS (regular OR irregular) at all since it is conversely completely devoid of opposed elements. Intuitively it seems like we should be able to bound this with two parallel planes, but I don't think it will be able to properly constrain a mating envelope. It could perhaps be type B however I'm not sure about the implications for that. Reading through a related post from some time ago (https://www.eng-tips.com/viewthread.cfm?qid=299827) seems to suggest it could not be regarded as a FOS - perhaps the thinking has changed since then.

Its interesting to note that there is one similar feature in the standard with figure 2-5 in conjunction with the dimension origin symbol and directly toleranced dimensions, this would seem to suggest that even without opposed elements it is a FOS of some sort however since the dimension origin symbol places additional constraint (ie: full/maximum contact with the indicated side is required) I believe that utilization of the dimension origin symbol is a totally different animal as a result and a similar conclusion cannot be made for standard directly toleranced dimensions.

Edit: Grammar

### RE: Feature Of Size definition

Andrew,

Any chance you could also post the section which defines regular vs. irregular FOS in the 2018 standard (I too am awaiting my copy)? It would be equivalent to 1.3.32 in the 2009 version - this would be most pertinent to the discussion at hand I think.

Side note - I'm interested to see what content was added before the envelope principle which was previously 2.7.1 and is now apparently 5.8.1 - 3 whole sections have been added, unless it was moved. Hopefully this added material has some value and doesn't just up the page count...

### RE: Feature Of Size definition

(OP)
Thanks everyone for the input!
Andrew, thank you for attaching Rule#1 from the new standard. Would you say it applies to the diameter 55? Diameter 55 would need to be considered Regular FOS for Rule#1 to be applicable. Fig. 5-9 in the attachment shows a regular feature of size with an interruption. I wonder how interrupted a feature can be and still be considered regular. I also join chez311 in the request for a post of the definition of Feature Of Size and its' types.

chez311, thanks for the link. It seems from that discussion that a feature such as width 52 in my sketch may not be considered a feature of size at all. I was planning to control that width for a central location by Position tolerance RFS to a primary datum axis derived from a shank at the extent of the part (not shown in the sketch because of the clipping). There would also be a rotational degree of freedom constrained by a flat that exists on the shank - a secondary datum feature. Do you think it would be a bad practice since the 52 dimension width will fail to "properly constrain a mating envelope"?

### RE: Feature Of Size definition

There is a discussion on linkedin and there are some opinions that E feature it is not an irregular feature of size.

“F" has been agreed that is not an irregular feature of size nor a regular one

“E” not so much agreement achieved.

Copy-paste

“E and F are not irregular features of size either. For these two be irregular features of size the mating envelope - two parallel planes - would need to able to close down on these faces. With these geometry closings parallel planes about the part surfaces would pitch ( or rotate) the part and would not actually close down on these faces.”

“E is an irregular feature of size. This is because we can define a mathematical envelope (2 parallel planes spaced at least 62 apart)”

Correction edit (mistakes in my original post). Hopefully now corrected

"E" from my imbedded picture resembles to yours 52 width feature, don’t they?

### RE: Feature Of Size definition

Sem,

I would say that your 55 diameter feature is less "interrupted" and more "incomplete" as I said it looks to be fully comprised of opposed elements - albeit just not fully 360deg around. This is why I said I could see the case for it being a regular FOS.

#### Quote (SemD220 8 Mar 19 06:28)

Do you think it would be a bad practice since the 52 dimension width will fail to "properly constrain a mating envelope"?
Yes I do - I would probably just forgo any possible argument/issues with inspection down the line and specify it as profile if possible.

Kedu,

Your copy-paste of that linkedin discussion highlights exactly what I mentioned in my original post when I said "Intuitively it seems like we should be able to bound this with two parallel planes, but I don't think it will be able to properly constrain a mating envelope" as I can see where both sides are coming from. That said, I lean towards it not being a FOS at all (regular or irregular).

#### Quote (Kedu 8 Mar 19 12:52)

“E is an irregular feature of size. This is because we can define a mathematical envelope (2 parallel planes spaced at least 62 apart)”
This is what I meant by "Intuitively it seems like we should be able to bound this with two parallel planes" as I think we can all imagine who two parallel planes could theoretically make up a boundary for this feature.

#### Quote (Kedu 8 Mar 19 12:52)

“E and F are not irregular features of size either. For these two be irregular features of size the mating envelope - two parallel planes - would need to able to close down on these faces. With these geometry closings parallel planes about the part surfaces would pitch ( or rotate) the part and would not actually close down on these faces.”
This is what I meant by "I don't think it will be able to properly constrain a mating envelope" since as the boundary closes down the feature will rotate as that boundary closes in, as it is not stable, having no opposed elements. I guess one could specify that the boundary closes in until it makes full/maximum contact with one or both sides (which is essentially what the origin dimension symbol does - perhaps a custom note could accomplish the same if really desired) but thats not really in the definition is it? I guess neither technically is the requirement for being "stable" - however it would be a logical conclusion I think since it is not stable there is no limit to where the contraction will stop, rendering it invalid in my mind. Even features/collections of features having no "directly" (180deg) opposed elements can still be stable as the boundary closes in - see Fig 4-35 in the 2009 version.

### RE: Feature Of Size definition

(OP)
Andrew,
Thanks for the attachment of the Feature Of Size definition from the new standard. I took notice that there is a definition of the term "Interruption" in the standard. It is a good sign because since it is defined, it is probably further addressed in other places in the standard - something that I felt was missing previously.

#### Quote (chez311)

I would probably just forgo any possible argument/issues with inspection down the line and specify it as profile if possible

I envy you :)
Our inspection does not know how to "digest" Profile tolerance. For the 2 flat surfaces 52mm apart, they would not have any issues at all if a horrible thing like 2X 26+-0.1 was specified between the axis and the surface. But they would probably raise questions about a basic 52 and a Profile call out referencing an axis datum and a clocking datum, because they think that Profile is intended mostly to control an accurate form (not location) of features with complex contours and always requires the use of CMMs and programming and what-not. Unfortunately, I don't know enough about metrology to convince anyone otherwise. But these are "my" special case problems, and I agree that ideally, Profile would be a better choice to avoid any "legality" issues.

Kedu, thanks for the reference, these are some interesting quotes from the linkedin discussion. "E" does resemble the feature from my case dimensioned by 52+-0.2.

"E and F are not irregular features of size either. For these two be irregular features of size the mating envelope - two parallel planes - would need to able to close down on these faces. With these geometry closings parallel planes about the part surfaces would pitch ( or rotate) the part and would not actually close down on these faces"

"F" is obviously not a Feature Of Size. As for "E", I'm less convinced by the quoted argument.
For this type of geometry, the mating envelope would have to be simulated as part of an inspection of Position control (or a center plane orientation control, for that matter). When simulated by physical gauging equipment, the part will have to be immobilized in a fixture with its' degrees of freedom constrained according to the referenced DRF. The specified DRF can prevent the rotation of the part during the simulation of the actual mating envelope, and there might not be an issue at all. Am I missing something?

### RE: Feature Of Size definition

Sem,

I should not have said there would be no issues. What I mean that I figured for a "by the book" legality standpoint it would be better and also more digestible by an inspection department. Perhaps thats not the case, but would they then have then no issue with you applying a size tolerance to the referenced feature?

To your second point, I agree that a RAME will be determinable for a position/orientation tolerance as long as it is sufficiently constrained to the DRF to prevent the unwanted movement. That said, a UAME will not be determinable and the concept of "size" will have no meaning.

### RE: Feature Of Size definition

(OP)
chez311,
If by "applying a size tolerance to the referenced feature", you mean 52+-0.2 (although this feature is not "referenced" as in datum references, so please clarify if you meant something else), they would probably not mind the size specification, as long as they manage to measure it. But I do understand that it is problematic; since according to what I've learned today it is probably not a feature of size, it should probably not be directly toleranced as well.

Speaking of that issue, I wonder if the new standard still shows direct +- toleranced dimensions on non-FOS like in fig. 2-4 in the 2009 version.

I understand your point about the difficulty to simulate the UAME for non-opposed "width" features.

### RE: Feature Of Size definition

(OP)
chez311, I just realized something that made me question our conclusion about undeterminable UAME for "E" type features:

#### Quote (chez311)

To your second point, I agree that a RAME will be determinable for a position/orientation tolerance

Actually, when you look for the axis or the center plane of a feature controlled by Position/Perpendicularity/Parallelism/Angularity RFS, you are looking for the UAME, not for the RAME. Think of it this way - checking the orientation of the center plane of the RAME is redundant, as the RAME is basically oriented to the referenced DRF. So, since the part is immobilized according to the DRF during the UAME simulation, I suppose that the unwanted movement is prevented after all. The thing to realize is that the part is constrained, but the simulated UAME is not constrained to the DRF, it merely follows the as-produced the feature. However, it can be simulated in a repeatable way. What do you think?

### RE: Feature Of Size definition

Sem_D220,

Datum reference frames are not involved in the determination of unrelated actual mating envelopes.

pylfrm

### RE: Feature Of Size definition

pylfrm -- read his post again. It's true that a DRF doesn't determine the UAME, but the UAME has to be derived for many geometric controls in order to compare that UAME to the DRF (such as positon or perp that he mentioned).

### RE: Feature Of Size definition

Belanger,

To be more specific, I disagree with this statement:

#### Quote (Sem_D220, 9 Mar 19 07:36)

So, since the part is immobilized according to the DRF during the UAME simulation, I suppose that the unwanted movement is prevented after all.

pylfrm

### RE: Feature Of Size definition

(OP)
Thank you, Belanger.

What I'm saying is that if a position control, for example, was applied to 52 size feature in my sketch referencing datums that constrain all rotational degrees of freedom for the part and 2 translational degrees of freedom perpendicular to a datum axis (which is designed nominally parallel to the center plane of the 52 size feature), the inspector would not experience any problems during simulation of the UAME of that feature. Again, the UAME is never constrained to any datums, but the part during UAME simulation is (in this case).

### RE: Feature Of Size definition

#### Quote (Sem D220)

the inspector would not experience any problems during simulation of the UAME of that feature. Again, the UAME is never constrained to any datums, but the part during UAME simulation is (in this case).

I'm not sure it matters how the part is physically held during simulation/measurement of the UAME - it doesn't change the fact that as pylfrm noted datum references are not involved in the creation of the UAME - by definition it is not constrained to any datums, only to the feature itself. Even if the part can be fully constrained this would only aid in the determination of the RAME, but would not prevent or limit the pitching/rotating of the UAME as it contracts around the feature.

I stand by what I said before that I do not believe a UAME is determinable for a feature of this type. If you disagree, could you maybe provide an example of what you think it would look like? Furthermore the concept of size I do not believe has any meaning either as it lacks any *opposed points (which in turn is also the reason a UAME cannot be created). Y14.5-2018 provides some clarity here which perhaps helps in some previously indeterminable cases (for example - if instead of 2x we had 3x offset planar features lacking *opposed points) but not in this one as it still requires the establishment of a UAME. As a side note, I am curious as to how this jives with either math standard (1994 or 2018) - my guess is it doesn't since per the discussion here (https://www.eng-tips.com/viewthread.cfm?qid=448687) it was commented that the new Y14.5.1-2018 is only fully compliant with Y14.5-2009...

*Edit: re-reading this I realize I lack the proper terminology here. Even replacing opposed with directly vs. indirectly opposed points seems equally ambiguous. Hopefully it can be easily visualized how 2 offset planar features of the type shown in the OP would be "unstable" as a boundary contracts around it, while adding a third offset feature would "stabilize" it. If that or my original meaning is not clear and a figure/example is needed someone please let me know.

### RE: Feature Of Size definition

Got it pylfrm. I haven't followed this thread carefully, and I just read things quickly.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

Even if the part can be fully constrained this would only aid in the determination of the RAME, but would not prevent or limit the pitching/rotating of the UAME as it contracts around the feature.

If the part is fully constrained, I hope we agree that the feature itself would not be able to pitch/rotate during simulation of the UAME (this is what I initially thought you were referring to).
The UAME would be able to rotate and translate (edit: and obviously contract) during the simulation until it becomes "a similar perfect feature(s) counterpart" which I interpret as 2 parallel planes making maximum possible contact with the high points of the feature, i.e. with both planar surfaces nominally 52mm apart. Once it complies to this definition/requirement, I think it is fully defined and determinable. What are your objections to that?

### RE: Feature Of Size definition

(OP)
Belanger, I think you actually understood my point pretty well originally, as I was talking in the context of UAME simulation during Position/Orientation controls - including the statement pylfrm quoted.

### RE: Feature Of Size definition

#### Quote (Sem D220 13 Mar 19 07:52)

I was talking in the context of UAME simulation during Position/Orientation controls
As far as I am aware of, there is only one definition of UAME which is wholly independent of what control is utilized, material condition/boundary modifiers, what the DRF looks like or if there is 0, 1, 2, 3, etc.. datum(s) called out - it does not matter, none of these things affect the resulting UAME.

#### Quote (Sem D220 13 Mar 19 06:45)

If the part is fully constrained, I hope we agree that the feature itself would not be able to pitch/rotate during simulation of the UAME
This is where we differ. It wouldn't be the feature/part itself which pitches/rotates/translates during simulation but the UAME boundary. The specified DRF has no impact on this regardless of whether the part is unconstrained or fully constrained.

#### Quote (Sem D220 13 Mar 19 06:45)

The UAME would be able to rotate and translate (edit: and obviously contract) during the simulation until it becomes "a similar perfect feature(s) counterpart" which I interpret as 2 parallel planes making maximum possible contact with the high points of the feature, i.e. with both planar surfaces nominally 52mm apart
The definition of the UAME has no requirement that there be maximum possible contact, or specify any importance of a nominal dimension. I think the driving part of the definition is that it is the boundary (similar perfect feature(s) counterpart) of "smallest size that can be contracted about an external feature(s) or largest size that can be expanded within an internal feature(s)" - ie: the UAME continues to contract/expand until its reached its smallest (minimum) or largest (maximum) size respectively. Perhaps someone else can support/refute this but I believe the reference to the fact that it "coincides with the surface(s) at the highest points" is not a requirement for maximum contact of any sort but instead a clarification/refinement to the requirement that the boundary exist wholly outside the material.

#### Quote (ASME Y14.5-2009)

1.3.25 Envelope, Actual Mating
envelope, actual mating: this envelope is outside the
material. A similar perfect feature(s) counterpart of
smallest size that can be contracted about an external
feature(s) or largest size that can be expanded within an
internal feature(s) so that it coincides with the surface(s)
at the highest points. Two types of actual mating envelopes
— unrelated and related — are described in paras.
1.3.25.1 and 1.3.25.2.

1.3.25.1 Unrelated Actual Mating Envelope.
unrelated actual mating envelope: a similar perfect feature(s)
counterpart expanded within an internal feature(s) or
contracted about an external feature(s), and not constrained
to any datum(s).

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

the UAME continues to contract/expand until it reached its smallest (minimum) or largest (maximum) size respectively.

The moment when it reaches it's smallest/largest size is the exact time when it contacts the surfaces at the highest points - there can be no other situation physically unless we constrain the UAME directly to some external reference, and we agree that we don't. And - in the same moment mentioned above, the UAME is constrained by the feature itself (which in turn is constrained by the DRF and is unable to rotate). Once again: the actual feature is constrained in movement by the DRF. The UAME is constrained by the feature when it reaches its' smallest size (for external feature) AND touches the feature on the highest points.

### RE: Feature Of Size definition

#### Quote (Sem D220 13 Mar 19 10:07)

the actual feature is constrained in movement by the DRF

#### Quote (Sem D220 13 Mar 19 10:07)

The UAME is constrained by the feature when it reaches its' smallest size (for external feature)

These two things are mutually exclusive. The DRF and accompanying constraint of the feature has no impact, directly or indirectly, on the definition/derivation of the UAME - the UAME remains the same if 0 DOF or 6 DOF are constrained. The way I read it, your statement suggests some indirect impact/constraint of the UAME resulting from the specified DRF, which is not the case.

### RE: Feature Of Size definition

(OP)
I was merely describing the physical conditions that are present during the UAME simulation. The bottom line and the message I wanted to convey is that there shouldn't be any rotation or pitching during the UAME simulation of the considered feature. The UAME can't rotate with the feature because the feature is constrained, and it can't rotate relative to the feature because there is only one way it can reach minimum parallel planes separation while contacting the surfaces of the feature.

Let me ask you this - let's call the two plane surfaces nominally 52mm apart "indirectly opposed" as you suggested. What advantage does a "directly opposed" width feature have over this feature? Why the UAME for a "directly opposed" feature can't rotate/pitch, whereas this one can't can?

### RE: Feature Of Size definition

#### Quote (Sem_D220, 13 Mar 19 12:07)

The UAME can't rotate with the feature because the feature is constrained, and it can't rotate relative to the feature because there is only one way it can reach minimum parallel planes separation while contacting the surfaces of the feature.

What are you suggesting defines a minimum separation? If your answer is still "maximum possible contact", please provide a definition for that term. Also, please explain the connection between what you suggest and the text of the standard.

pylfrm

### RE: Feature Of Size definition

(OP)
"Minimum separation" means the minimum achievable distance between parallel planes as they contact the surfaces of the feature. This condition is achieved when the maximum contact exists between the UAME simulator and the feature, although the definition of the term Actual Mating Envelope (Y14.5 2009 para. 1.3.25) in the standard only mentions contact on the highest points without explicitly requiring "maximum" contact. "Maximum contact" is mentioned in a paragraph that describes a closely related concept - simulating a datum center plane of a feature of size referenced RMB:

#### Quote (ASME Y14.5 para. 4.11.4)

(b) Primary Datum Feature: Width RMB. The datum is the center plane of the datum feature simulator of the datum feature. The datum feature simulator (or unrelated actual mating envelope) is two parallel planes at minimum separation (for an external feature) or maximum separation (for an internal feature) that makes maximum possible contact with the corresponding surfaces of the datum feature. See Figs. 4-3, illustration (b)

As you can see, the definition itself includes a clarification in parenthesis that the simulation of a datum center plane is essentially establishment of an unrelated actual mating envelope and both "minimum separation" and "maximum possible contact" are mentioned in the same sentence. Analogical descriptions appear for cylindrical and spherical datum features in the same paragraph.

### RE: Feature Of Size definition

Sem D220,

I can see what you're getting at, with the idea of defining the UAME in terms of maximum contact between the feature and a perfect-form counterpart. I agree that this would be a better way to define the UAME. The current definition was based on simple well-behaved features of size like fully opposed cylinders and widths. For these features, it happens that the size of the envelope is maximized (or minimized) when maximum contact is achieved. So it worked well enough to define the UAME in terms of the size of the simulator (or the separation of the planes in the case of the width feature). When we start looking at less ideal features like the unopposed planar surfaces or unopposed partial cylinders, maximizing (or minimizing) the size of the simulator doesn't work - because the important thing is the maximum contact.

Y14.5 has introduced the idea of maximum contact as you have pointed out, but there is still a ways to go. First, the term "maximum contact" is used in certain definitions (see Fig. 4-31 (a) in Y14.5-2009) but the meaning of the term is not defined.

I would say that Y14.5.1 will lead the way with improvements to the definition of maximum contact.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

### RE: Feature Of Size definition

(OP)
Thank you for the input axym.
You said:

"When we start looking at less ideal features like the unopposed planar surfaces or unopposed partial cylinders, maximizing (or minimizing) the size of the simulator doesn't work - because the important thing is the maximum contact."

If we deal with the case of the 52mm unopposed width from the sketch I posted, ignoring the addition of "maximum contact" condition at simulation and only base the verification process on the pure definitions of AME and UAME specifically in paragraphs 1.3.25 and 1.3.25.1, do you think that there might be a case where bringing the simulator planes to minimum separation about the feature will not result in repetitive and determinable result? If yes, is it any different with simple opposed width features? The reason I'm asking is that after examining it more closely, I no longer stand by my previous statement that there can only be one way to bring the simulator to a minimum separation about the feature. For example, where both surfaces of the feature are produced slightly convex, there could theoretically be 2 different orientations at which the UAME simulator can contract to the same size of minimum separation (maximum separation for this case would be when the simulator contacts the peaks at the central areas of the slightly convex surfaces). But - and this is the important point - the same issue can easily occur for regular opposed width features of size. So, I have yet to be convinced that the unopposed geometry can be the cause for unstable UAME simulation and consequently disqualify this type of features from being considered features of size.

The only essential difference I see between this type of features and regular features of size is the lack of an "Actual Local Size" per para. 1.3.54 at Y14.5-2009. But, the existence of actual local sizes is not a requirement that a feature must conform to in order to be considered a feature of size per the definitions in the subparagraphs of 1.3.32. If the UAME is determinable, and I am yet to be convinced that it's not, why feature-of-size related applications such as Position control can't be utilized for such features?

### RE: Feature Of Size definition

Sem_D220,

Maybe this will help:
https://files.engineering.com/getfile.aspx?folder=...

Shortly speaking, because the lower feature doesn't have opposed points, a gage simulating actual mating envelope of that feature is able to contract to the much smaller size than the shortest distance between two perfectly flat (not even slightly convex) faces constituting the considered feature. As chez311 mentioned in one of his previous replies, if instead of 2 there were 3 offset (at distance 52) planar features lacking opposed points (like for example datum feature A in fig. 4-33 in Y14.5-2009), the contraction would have to stop at 52, similar to the case of regular FOS shown in the upper portion of the attached graphic.

### RE: Feature Of Size definition

Sem,

My apologies. I was running out of ways to describe what I meant purely with words and did not have time yesterday to come up with a few cases and put them into figures. Hopefully the ones I have now made will communicate what I am trying to say and not confuse the issue further.

First off I want to reiterate/address something which I think is important and sort of preclude any discussion of maximum contact/UAME definition.

Your main premise starting with your post on (9 Mar 19 07:36) was that since the part could be fully constrained to a DRF during simulation, a UAME could be defined. I apologize for beating a dead horse but I think it bears repeating - it does not matter if the part is constrained, partially constrained, or unconstrained - the UAME definition will not change. If it did, it would violate the definition of UAME which does not involve datum references. If you want to attack it from a different angle by attempting to define maximum contact instead then I that is a different matter - I believe I still disagree, but at least I can see the merits of the argument.

See my attached figure for three cases.
Case #1 - offset faces are flat and parallel
Case #2 - offset faces are flat and non-parallel
Case #3 - offset faces are convex

In case #1, my assertion that the UAME boundary can progress without limit (1B*) so there is no "minimum size" and is therefore indeterminate. I understand your point about maximum contact but it is not in the definition of UAME and regardless is not well defined (if at all).

My case #2 further highlights this point. Where is maximum contact in your opinion? Is it 2B* or 2C*? Each has the same amount of contact, however 2C has a smaller envelope. As with case #1 there is no limit to the progression (and no minimum size limit to the UAME) so I say the UAME is still indeterminate.

For my case #3 I assert still that there is no limit to progression, now with no definable "maximum contact" as the UAME progresses the amount of contact never changes.

#### Quote (Sem D220 14 Mar 19 17:46)

For example, where both surfaces of the feature are produced slightly convex, there could theoretically be 2 different orientations at which the UAME simulator can contract to the same size of minimum separation (maximum separation for this case would be when the simulator contacts the peaks at the central areas of the slightly convex surfaces)
This would be my case #3. I believe there is no limit to progression and not 2 different orientations as you suggest.

#### Quote (Sem D220 14 Mar 19 17:46)

But - and this is the important point - the same issue can easily occur for regular opposed width features of size.
**

I'm actually not convinced this is the case. I had the same thought however after doing some quick modelling/sketching in CAD I explored some options such as a barrel-shaped convex feature with opposed points and there was always a single defined minimum separation of the UAME. If you can develop a figure which shows otherwise could you share it?

*Edit: a few incorrect references to my own figures

**Edit2: deleted some repeated portions of a quote

### RE: Feature Of Size definition

I saw a few replies as I was in the process of putting together this post. I also made a few edits - I hastily hit submit.

Sem - I apologize if you have shifted your thinking or changed your position in regards to the points I address. If so please kindly correct me where I may have missed the mark.

pmarc - I hope my figure is not too redundant, I read your post and I thought it was still worth sharing. You illustrate the point I was trying to make yesterday but did not have the time to put together figures to communicate clearly. Thank you.

### RE: Feature Of Size definition

(OP)
Thank you pmarc and chez311 for the clear figures.
I understand the points you convey, but I still suspect that similar issues (of unrestricted contraction) may occur with regular opposed features of size, given that we take "maximum contact" out of the equation. In 2 or 3 days from now when I have access to CAD again I will check this and update, and if what I imagine is correct, I will post the relevant figures.

#### Quote (chez311)

Your main premise starting with your post on (9 Mar 19 07:36) was that since the part could be fully constrained to a DRF during simulation, a UAME could be defined. I apologize for beating a dead horse but I think it bears repeating - it does not matter if the part is constrained, partially constrained, or unconstrained - the UAME definition will not change. If it did, it would violate the definition of UAME which does not involve datum references

chez311, I feel it's time this misunderstanding gets clarified. You are not beating a dead horse, you are preaching to the choir
In the post you refer to from 9 Mar 19 07:36 I clearly said:

"The thing to realize is that the part is constrained, but the simulated UAME is not constrained to the DRF, it merely follows the as-produced feature"

On 13 Mar I repeated:

"Again, the UAME is never constrained to any datums, but the part during UAME simulation is (in this case)."

I believe I said similar things in a few more occasions in this thread later. The reason I brought up the part being constrained by the DRF in this discussion is that it seemed to me that the main support for the claim about unstable UAME simulation was originally the possible movement of the part/feature during contraction of the UAME simulator. This idea was clearly described in the quote supplied by Kedu from the LinkedIn discussion:

"E and F are not irregular features of size either. For these two be irregular features of size the mating envelope - two parallel planes - would need to able to close down on these faces. With these geometry closings parallel planes about the part surfaces would pitch ( or rotate) the part and would not actually close down on these faces.”

In response to this quote you said:

#### Quote (chez311, 8 Mar 19 14:33 )

This is what I meant by "I don't think it will be able to properly constrain a mating envelope" since as the boundary closes down the feature will rotate as that boundary closes in, as it is not stable, having no opposed elements.

It appears that later, you changed your perspective and approached the UAME instability issue differently:

#### Quote (chez311, 13 Mar 19 09:39 )

This is where we differ It wouldn't be the feature/part itself which pitches/rotates/translates during simulation but the UAME boundary.

Prior to that restatement, I was already commenting on the original claim and my point was, that if the part/feature is constrained per the DRF during the inspection which involves simulating a UAME, the issue described (movement of the part) is irrelevant. To add to that I'd like to say now that as much as the UAME simulation should be independent of the referenced datums by definition, it should also be independent of phenomena such as uncontrolled rocking/wobbling/pitching/rotation of the part as the result of the forces acting on it during the inspection.

### RE: Feature Of Size definition

Hi All,

pmarc's figures illustrate exactly what I was thinking. The upper example with the I-beam shape is the type of well behaved geometry that Y14.5's minimum-separation definition works well on. The unopposed geometry in the lower example makes the definition fail. The B figure represents what I was thinking with the maximum contact.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

### RE: Feature Of Size definition

(OP)
pmarc, axym,

Illustration C for the unopposed geometry in pmarc's feature figure shows the part tilted relative to its' previous condition at illustration A and B. To me this suggests that the failure in UAME simulation happens because of a degree of freedom that the part has. As I mentioned I do realize that no datum constraints should affect the UAME simulation result, but the simulation should neither be affected by any degrees of freedom that the part has and the forces applied to it during an inspection. Am I wrong?

I'm aware that essentially it doesn't matter if the part is rotated relative to the simulator or the simulator is rotated relative to the part. I could tilt my head to the side when looking at illustration C and view it as if the simulator was rotated relative to the part and not the other way around. Nevertheless, (edit: no matter what rotates and what stays at constant orientation in the part - simulator interaction), I do think it is a point worth looking into because there are procedures that are intended to prevent the effect of forces in inspection and perhaps the problem can potentially be avoided altogether. I believe there is a paragraph in ASME Y14.43 dealing with the issue.

As promised I will try to make figures of my own in the next few days if my suspicion that similar issues may occur with regular opposed features will turn out justified.

### RE: Feature Of Size definition

Sem,

I caught your comments about the UAME not being constrained to any DRF, which made your assertions that inclusion of a DRF and constraint of the part would enable the derivation of a mating envelope all the more confusing. I did not catch my own change in wording - for me the frame of reference was inconsequential as long as I kept in mind that inclusion or removal of datum references would have no effect on establishment of the UAME. Your statement that "essentially it doesn't matter if the part is rotated relative to the simulator or the simulator is rotated relative to the part" is what my thought process was the whole time. I should have been more consistent in my wording and/or caught that and clarified myself.

#### Quote (Sem D220 15 Mar 19 09:00)

As I mentioned I do realize that no datum constraints should affect the UAME simulation result, but the simulation should neither be affected by any degrees of freedom that the part has and the forces applied to it during an inspection. Am I wrong?
I think I agree with you here. This would be the physical reality of the theoretical specification that datum references are not involved in establishment of the UAME. That is as long as you mean forces/constraints as it applies to any datum features on the part.

In regards to maximum contact, I have always had a problem with the fact that this term was continually utilized in the standard yet never defined. Without this we are continually left to wrestle with what it means in different circumstances - in lieu of this definition I have typically taken it to mean that it progresses until it cannot progress any further. Any other definition creates constraints which may stop progression before that point, this could be acceptable if it is clearly specified somehow, but not in my mind going solely by the content in the 2009 standard. By a brief skim of the 2018 standard it does not look like much clarity is provided - though I will have to take Evan's comment into account and do a more thorough read through of the Y14.5.1 draft and see what clarity may have been imparted.

I am interested to see what you come up with for a purportedly more "well behaved" FOS (as Evan called it) having opposed points. I can perhaps see situations where there might be multiple solutions (probably 2, maybe more) but none analogous to the one in your initial post which I believe has no solution.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

I caught your comments about the UAME not being constrained to any DRF, which made your assertions that inclusion of a DRF and constraint of the part would enable the derivation of a mating envelope all the more confusing.

chez311,
I'm sorry if I put my ideas into words in a confusing way. But the message I try to express about the effect of DRF is consistent throughout. The DRF is not part of the UAME definition. But, where it participates in the inspection process, it may prevent the phenomenon that was argued (not by me) to be the cause of infeasible UAME simulation. This phenomenon is the uncontrolled movement of the part as a result of forces that the simulator exerts on the part. If we accept that this is a legitimate argument, then the DRF being part of the picture during UAME simulation can be brought up as a counterexample. In that specific context, the fact that it is the part/feature which was pointed to as the unstable (rotating/pitching) element in the process matters - obviously because the DRF affects the part, but never the UAME . Again, all of the above is relevant only if we look at the "rotating part" issue as a legitimate reason to dismiss unopposed features as non-features-of-size.

The additional point I was suggesting to examine in my recent posts, is that the relative movement between the part and the simulator is irrelevant altogether to the whole concept of the definition of the feature of size. That is because the presence of degrees of freedom and the ability of forces during inspection to cause movement should not affect the UAME, as all these are not part of the UAME definition, just as datums are not. In this specific context, it doesn't matter if the part is rotated relative to the UAME or the UAME is rotated relative to the part. I suggest questioning the relevance of relative movement generally to the subject.

Hope this clears up some of he confusion that was apparently caused by my previous posts. And I apologize for causing that confusion.

### RE: Feature Of Size definition

#### Quote (Sem D220 15 Mar 19 15:43)

The DRF is not part of the UAME definition. But, where it participates in the inspection process, it may prevent the phenomenon that was argued (not by me) to be the cause of infeasible UAME simulation. This phenomenon is the uncontrolled movement of the part as a result of forces that the simulator exerts on the part. If we accept that this is a legitimate argument, then the DRF being part of the picture during UAME simulation can be brought up as a counterexample.
I do not agree with this. I accept that there are realities in the physical world with inspection that make holding precisely to the theoretical concepts outlined in this (and other) standards difficult - however the setup must account for that as much for that as possible. If the UAME must be simulated (which it does not always - say for most MMC/MMB situations) then it might require an additional setup/gauge/fixture - the datum features should not be allowed to limit any DOF in this instance, within reason of course I accept there could be limitations to what is possible.

#### Quote (Sem D220 15 Mar 19 15:43)

The additional point I was suggesting to examine in my recent posts, is that the relative movement between the part and the simulator is irrelevant altogether to the whole concept of the definition of the feature of size. That is because the presence of degrees of freedom and the ability of forces during inspection to cause movement should not affect the UAME, as all these are not part of the UAME definition, just as datums are not.
I'm not really sure how to address this, or what this adds to the discussion about establishing a UAME for features of this type. I'm going to try my best to answer it but I'm not sure I'm going to succeed. Perhaps someone else can pick up the slack where I may be lacking. Here goes:

Relative movement of the simulator and feature and the forces exerted as a result (though hopefully very small in order to minimize deflection) is just a result of the physical realities of simulating a theoretical boundary (UAME) with a physical part (simulator/gauge), pursuant to the limitations I noted above. I guess in theory this boundary would be coincident with the surface and is inseparable - ie: does not have to be brought into contact and instead coexists with the surface of the feature, that is if it even exists (as in the case where there is no solution). Relative movement is also an easier way for us as humans to visualize some of these situations (I know it is for me). I think I understand what you are saying, essentially if there is a solution which achieves "maximum possible contact" with a feature then this could define a boundary even if an imbalance of forces would exist with a physical gauge/simulator (ie: is not a stable equilibrium) - however "maximum possible contact" is not defined in the standard so I'm not sure where you would go with that. You could come up with your own specification for this if desired or call out a particular fitting routine.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

If the UAME must be simulated (which it does not always - say for most MMC/MMB situations) then it might require an additional setup/gauge/fixture - the datum features should not be allowed to limit any DOF in this instance

What do you suggest to do for simulation of a UAME of an inspected feature as part of position RFS or center plane/axis orientation RFS control? Without establishing the DRF the tolerance zone is undefined. How do you suggest to verify those tolerances without limiting the degrees of freedom of the part? Maybe I misunderstood your statement, if so please clarify.

Regarding the second point - perhaps there is another way to apply an unambiguous UAME solution, without demanding "maximum contact" which is apparently not well defined and not part of the definition. Also considering that we insist on not simulating a controlled feature's UAME the same way we derive datums from datum features of size (para. 4.11.4 which requires maximum possible contact). The other way I'm suggesting is looking at the term "a similar perfect feature(s) counterpart" which is part of the definition and appears at para. 1.3.26. How should the term "similar perfect feature counterpart" interpreted and can it imply a definitive single solution? Looking for more details I found out that Alex Krulikowski defines it as a "boundary that is the perfect inverse of the feature" (Fundamentals of Geometric Dimensioning and Tolerancing). In the same source, he also writes that an actual mating envelope is "a similar perfect feature counterpart that would surround the high points of a feature of size". From this and from the definition in the standard I get the impression that contraction to the minimum size about an external feature or expansion to the maximum size within an internal feature is not the only, and perhaps not even the main role of the UAME. More specifically, there is a limit until which the UAME simulator of any feature can contract/expand without becoming noncompliant to the specification of being a similar perfect feature counterpart - which actually means staying as adjacent as possible ("perfect inverse") to the feature.

If my personal opinion matters, I still say that the most rational way to approach this is to simply follow paragraph 4.11.4 for every case where a UAME should be simulated. I may not have a solid argument here to back it up, but my logic tells me that there is no reason to treat the derivation of an axis or a center plane of a feature that should be controlled for location or orientation differently than the derivation of a datum axis or center plane. I'd like quote part of this paragraph once again:

"(b) Primary Datum Feature: Width RMB. The datum is the center plane of the datum feature simulator of the datum feature. The datum feature simulator (or unrelated actual mating envelope) is two parallel planes at minimum separation (for an external feature) or maximum separation (for an internal feature) that makes maximum possible contact with the corresponding surfaces of the datum feature. See Figs. 4-3, illustration (b);
4-13; and 4-14."

### RE: Feature Of Size definition

This is just my opinion, but this thread is a great example of why the standard should show (for example in an appendix) some examples of features that are not regular and irregular features of size. I realize it is impossible to show all cases one can ever imagine, but showing some of the most common scenarios definitely wouldn't hurt.

### RE: Feature Of Size definition

(OP)
The image shows 2 cases comparable with Case #2 and Case #3 in the figure by chez311. The same issues that were associated with non-opposed geometry occur for features designed to be "well behaved" opposed feature of size. Left sides shows a feature produced with slightly convex faces, and 2 possible candidate UAMEs at different orientations. This is the "well behaved" equivalent to Case #3. Right side shows a feature with the faces produced slightly non-parallel. This is the "well behaved" equivalent to case #2.

### RE: Feature Of Size definition

Sem_D220,

If we, for example, take you picture on the right:
1. What makes you think that the 19.49 envelope is really the UAME of the as-produced 20 feature?
2. More important, is further contraction of the 19.42 envelope possible?

### RE: Feature Of Size definition

(OP)
pmarc,
I never said that 19.49 is the correct size of the UAME.

### RE: Feature Of Size definition

Sem_D220,

1. Ok. I simply wanted to make sure that you did not consider the 19.49 envelope good candidate for UAME.

2. I personally do not think that the '19 or less' envelope is good candidate for UAME too. It is basically because the as-produced 20 feature is not fully constrained/immobilized relative to that envelope. To achieve full immobilization, the envelope would have to contact the feature in at least 3 points in the shown view (2 on one face and 1 on second face). For the proposed as-produced geometry, it is the 19.42 and the 19.49 envelopes that satisfy this condition, but because the 19.42 envelope is smaller of the two, it is the correct UAME of the feature.

This is how I understand maximum possible contact with the corresponding surfaces of a planar (two-parallel-planes) feature of size.

### RE: Feature Of Size definition

(OP)
pmarc,
Just to clarify in case someone may misunderstand, the number 19 ("or less") was randomly chosen to be shown as an example. If we go purely by the UAME definition as was advocated by numerous responses in this thread then "maximum contact" is not part of the considerations and therefore any value below 19.49 19.42 and above the width of the feature (for example - 5) would be a legitimate candidate too, and could be chosen as an example for the answer on your question: "is further contraction of the 19.42 envelope possible?"

Per my understanding, not even the 3 point contact you described would save the situation, because technically all three points of contact can be located on the edges intersecting the top/bottom faces and the 2 left/right surfaces. For example - 2 points on the edge between top & right and one point on the edge between bottom & left, for the 19 UAME shown. I agree that this kind of contact would not immobilize the feature relative to the envelope, as you say. The same is true for the <52 envelope in illustration C for the unopposed feature which you posted above. Both envelopes are equally appropriate or inappropriate.

For immobilizing the feature, the unrelated actual mating envelope has to do more than just contracting to the smallest size about the external feature until further contraction is impossible. It has to be sufficiently adjacent to the feature as well. Whether the faces the feature consists of are opposed or non-opposed doesn't seem to matter in this context.

### RE: Feature Of Size definition

#### Quote (Sem_D220)

Per my understanding, not even the 3 point contact you described would save the situation, because technically all three points of contact can be located on the edges intersecting the top/bottom faces and the 2 left/right surfaces. For example - 2 points on the edge between top & right and one point on the edge between bottom & left, for the 19 UAME shown.

If you read my last reply again, you will notice that I said that at least 3 points of contact would be required in the shown view. What you described, as quoted above, doesn't satisfy this condition.

#### Quote (Sem_D220)

The same is true for the <52 envelope in illustration C for the unopposed feature which you posted above. Both envelopes are equally appropriate or inappropriate.

I disagree with that for the reasons I already explained in my previous comments.

### RE: Feature Of Size definition

(OP)
pmarc, I'm sorry for missing that you said that all 3 points of contact should be in the shown view. I hope you don't mind answering on these 3 questions:

1. We are dealing with a 3-dimensional feature and a 3-dimensional UAME, how do the 3 points of contact in a shown view prevent the same problem (of the UAME and the part not being constrained to each other, and the unrestricted contraction of the UAME simulator) from occurring with rotation around another axis in a way that can be visualized on another drawing projection?

2. How is the "3 points of contact" concept implied by the UAME definition in the standard? The background to this question is that I was told that "maximum possible contact" shouldn't be required during UAME simulation as it isn't part of the definition. And since it isn't required, UAME simulation might fail - but only for unopposed features.

3. Considering that the issue described in the first question can somehow be resolved or there is no issue at all and it is me who fails to see it (quite probable), why can't the same concept be implemented on unopposed features such as in the second row of the illustration you posted above? If this is already covered in your previous posts, I apologize for missing this too.

### RE: Feature Of Size definition

1 and 2. The concept of 3 points of contact in shown view was introduced just to picture that the feature should be immobilized relative to its true UAME. And of course because we are talking 3D, the same logic applies in the side view (I thought this did not have to be explicitly stated).

The key thing here is that as long as in any of the two views the 3 points of contact are seen as 2 points, there will be an instability of the feature relative to the envelope. In other words, the envelope will not be such that it "coincides with the surface(s) at highest points", as given in the UAME definition.

3. Yes, it has been already covered by me. It is not the concept of 3 points of contact that really matters in case of a feature having non-opposed surfaces. It is the idea that there is no way to fully immobilize the feature relative to the contracting envelope because there is nothing that physically stops the contraction of the envelope.

As pointed out by chez311, physical reality is what should really matter here. This should also drive the classification of a feature as FOS or non-FOS. Definitions in the standard should go hand in hand with the reality, but, as it was already mentioned, they unfortunately do not always work well for cases other than "well behaved" geometries produced as well behaved shapes.

### RE: Feature Of Size definition

(OP)
pmarc,
Now that I realize that the 3 points idea is not view dependent, I think this is the definition for "maximum possible contact" that is missing in the standard as was pointed out by axym, chez311 and possibly others. Perhaps such definition is not required and as you pointed out - 3 points of contact is the true meaning of the sentence "coincides with the surface(s) at highest points" and this specification should simply not be neglected. That is because as I've shown in the figure from 17 Mar 19 10:11, if this specification is not followed, even the UAME for a simple feature designed as a "well behaved" geometry, and possibly produced within all tolerances, might not have a solution.

#### Quote (pmarc)

It is not the concept of 3 points of contact that really matters in case of a feature having non-opposed surfaces. It is the idea that there is no way to fully immobilize the feature relative to the contracting envelope because there is nothing that physically stops the contraction of the envelope.

I'm sorry but I am still not getting it.
I have clearly shown that without the "maximum possible contact" or if you prefer "the 3 points of contact method", the statement "nothing that physically stops the contraction of the envelope" is equally relevant to simple "well behaved" features of size. Why would the 3 points method prevent the failure of UAME size < 19.42 from my figure from 17 Mar 19 10:11, but not the failure of UAME size < 52 from your figure from 14 Mar 19 18:23?

Edit: grammar correction

### RE: Feature Of Size definition

#### Quote (Sem D220 15 Mar 19 22:16)

What do you suggest to do for simulation of a UAME of an inspected feature as part of position RFS or center plane/axis orientation RFS control?
As for a physical gauge, I would really defer to someone with more experience in gauging - however I do know that this is one of the main reasons that gauges for RFS position are much more expensive. The first thing that comes to mind though is an expanding pin - this could be utilized to simulate the UAME of an RMB primary datum feature and could touch/constrain other features and DOF as long as it was expanded within the primary datum feature first. Another could be a free floating expanding pin which is inserted and expanded within a hole and does not contact any other features - this could be utilized to find the size of the UAME (or a series of solid plug/pin gauge could be inserted until one is found which just fits - this would provide a similar function). As far as how to utilize this simulated UAME to gauge RFS position I would let someone more experienced than I answer that.

#### Quote (Sem D220 15 Mar 19 22:16)

Without establishing the DRF the tolerance zone is undefined. How do you suggest to verify those tolerances without limiting the degrees of freedom of the part?
UAME =/= tolerance zone. The UAME must fall within the established tolerance zone - or in the case of a primary datum feature, it establishes the DRF from which other features are derived.

#### Quote (Sem D220 15 Mar 19 22:16)

If my personal opinion matters, I still say that the most rational way to approach this is to simply follow paragraph 4.11.4 for every case where a UAME should be simulated.
I do not see how this helps clarify matters considering the additional term "maximum possible contact" is still not defined within the bounds of the standard. As far as I'm concerned, the 3 points of contact is just a result of maximum expansion/contraction/progression of the boundary to reach its largest (for an internal feature) or smallest (for an external feature) size. Even if 3 points of contact can be established before then (in your example - the 19.49 boundary) it is not the UAME unless it has reached its smallest size (the 19.42 boundary).

In regards to your two features, for the second one I agree with pmarc that the correct UAME is unambiguously 19.42 - your alternative "19 or less" is not valid. If you consider contraction of the boundary as it progresses towards its minimum separation, 19.42 is the only solution - to achieve a smaller boundary (your "19 or less") it is no longer only contracting but also rotating arbitrarily away from the surfaces being simulated.* Conversely this would also happen if you were starting contraction with the part at an excessive angle such that it closed down on the longer 2x faces instead of the shorter 2x faces in question.

In regards to your first figure, I was intrigued as to why I did not see that on my initial, admittedly quick, look at the geometry. It looks like with features of a high length to width ratio as you have shown, with relatively loose tolerances (not unreasonably so - but definitely not very tight) on the size/form an unstable situation could arise. If I have some time I might come up with a formula to show the relationship, but in the meantime the below figure shows that in order to create the situation you showed it requires the center of the arcs which make up the barrel shape to be at some separation from each other, which also requires a relatively small radius on each side resulting in a relatively large size/form deviation. If these arcs are changed so that their centers are coincident then the feature essentially now becomes round and there is no minimum as the separation becomes the diameter and is the same at every point. If these arc centers cross the centerline then there is now a minimum at the apex of each arc and the feature now becomes "well behaved" - as the size/form tolerance gets tighter this would only become more true. I don't know how often one might encounter this as it requires a specific set of conditions, and I do not know exactly how this would be handled in regards to establishing a UAME - however I do not think it is perfectly analogous to your OP example as they are not "unstable" in the same sense - ie: I think you would find that if you contracted a physical boundary about my case #2 below I think you would find it relatively stable, despite having no minimum. The same could not be said for your OP with offset planar features.

Note the difference in the size/form deviation of each feature (.462 / .318 / .310 respectively). As I stated if this is minimized by a tighter tolerance, this instability and lack of defined singular minimum disappears.

*I know I stated in my reply (15 Mar 19 18:19) that theoretically these boundaries would coexist with the surface and maybe downplayed the physical realities, based on pmarc's replies and thinking about it some more I think I was on the right track initially by focusing on the physical realities of simulation.

**Edit - made the figure more readable, I realize that the dimensions may have been hard to read.

### RE: Feature Of Size definition

(OP)
chez311, I'm not experienced in gaging either, but I think that the idea of checking position RFS or center plane orientation RFS of a feature of size is simulating the UAME while the part is constrained at the datum features at the fixture, and its' degrees of freedom are constrained as implied by the DRF specified in the drawing. When the UAME is simulated under these conditions, the derived feature axis or center plane (the product of the UAME simulation) is checked for fitting within the tolerance zone, which can only be determined based on the datum feature simulators.

#### Quote (chez311)

If you consider contraction of the boundary as it progresses towards its minimum separation, 19.42 is the only solution - to achieve a smaller boundary (your "19 or less") it is no longer only contracting but also rotating arbitrarily away from the surfaces being simulated

How would you comment on the suggestion that the boundary of 39.030 for Case #2 of the unopposed feature from your post at 14 Mar 19 18:46 is also the only solution for that case? 47.444 is not the UAME size as it is possible to contract the envelope further without losing proper contact relationship with the feature. Less than 39.030 - unacceptable rotation and loss of contact occurs just like with the "simple" opposed feature. What is the essential difference?

#### Quote (chez311)

I think you would find that if you contracted a physical boundary about my case #2 below I think you would find it relatively stable, despite having no minimum.

I wouldn't call it stable because even though the size of the UAME stays constant the orientation of the UAME (and therefore the orientation of the feature center plane) can differ from one simulation to another. If the purpose is to check if it falls within a tolerance zone or not, there is going to be trouble.

### RE: Feature Of Size definition

#### Quote (Sem D220 18 Mar 19 17:51)

but I think that the idea of checking position RFS or center plane orientation RFS of a feature of size is simulating the UAME while the part is constrained at the datum features at the fixture, and its' degrees of freedom are constrained as implied by the DRF specified in the drawing.
This would no longer be the UAME. It would be the RAME. I'm not really sure how one could say otherwise - its right there in the definition of both envelopes.

#### Quote (Sem D220 18 Mar 19 17:51)

How would you comment on the suggestion that the boundary of 39.030 for Case #2 of the unopposed feature from your post at 14 Mar 19 18:46 is also the only solution for that case? 47.444 is not the UAME size as it is possible to contract the envelope further without losing proper contact relationship with the feature. Less than 39.030 - unacceptable rotation and loss of contact occurs just like with the "simple" opposed feature. What is the essential difference?
I would say as I have from the beginning - the feature in that post is not a FOS and has no determinable UAME as a result of the unlimited contraction. The multiple cases shown were to challenge your assertion about maximum possible contact - none of them are a FOS. This is not the same situation as your "19 or less" (which I have already said is invalid). Take two physical parts of the types shown and contract a boundary around them, the non-opposed feature (52 width) will allow unlimited contraction - the opposed feature (20 width) will stop contraction at 19.42.

#### Quote (Sem D220 18 Mar 19 17:51)

If the purpose is to check if it falls within a tolerance zone or not, there is going to be trouble.
I agree. I was saying it is not perfectly analogous to your OP, not that there was no issue - if you contracted a boundary around it, it would behave like any other round feature, it would stop contraction at a finite point (20). You are correct, this could have multiple orientations depending on how it was oriented in the simulator - which is why I conceeded that I wasn't sure how this would be handled to determine a UAME. I assume some sort of fitting routine to minimize separation between the boundary and the surface if this kind of variation could be expected - as I said it requires a very specific set of conditions and I doubt this is something that would typically be an issue. But I could be wrong.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

This would no longer be the UAME. It would be the RAME. I'm not really sure how one could say otherwise - its right there in the definition of both envelopes.

The mention of the unrelated actual mating envelope and not of the related actual mating envelope in the description under "means this" of fig. 7-65 is not a mistake. Note that the part, and consequently, the slot which is controlled for position, as well as the tolerance zone, are constrained to the |A|B| DRF. At inspection, 5 degrees of freedom of the feature are going to be constrained. As I said several times, the DRF constrains the degrees of freedom of the part/feature, but not of the UAME. To comprehend this, you can imagine being in a room where an object is screwed to the floor at a fixed distance from the walls. This object is the considered feature and your hands act as the UAME simulator. You are free to move your hands to all directions and rotate them as you wish, which is what the definition of the UAME prescribes. This is not in contradiction with fact that the degrees of freedom of the object you are going to touch are constrained, and with the fact that the whole process is only meaningful if performed inside the room - which is your datum reference frame. Unless you use excessive force, the object you are inspecting will not translate or rotate in the directions at which the constrained degrees of freedom prevent movement. This is why I brought up the DRF as a subject for consideration after it was stated that for unopposed features, the simulation process will result in movement (rotation) of the part and inability to simulate the UAME as a result of that movement.

#### Quote (chez311)

the opposed feature (20 width) will stop contraction at 19.42.

That is incorrect. If you simulate the UAME by a simulator such as a vise which is not attached to anything and is free to translate and rotate in all directions, if you keep tightening the jaws when the separation distance reaches the size of 19.42, the vise will not stay in equilibrium but will keep rotating in the direction which allows further contraction of the distance between the jaws. This is the physical behavior that will make the UAME simulator contract from 19.42 to 19 and further below, until finally closing on the width of the feature instead of its' height.

As for the scenario of the feature with convex faces, I think that there is very little benefit that can be gained from comparing which one is more ambiguous - the opposed case or the unopposed case. Both require a very specific set of conditions, not just the opposed one. I wouldn't try to predict at which case the failure is more probable. The fact that the opposed geometry doesn't guarantee a definitive UAME simulation result (or a solid physical barrier that will stop the contraction of the envelope simulator), is enough in order to call into question the differentiation between opposed and unopposed geometries in the context of the definition of feature of size.

### RE: Feature Of Size definition

Sem_D220,

"Take two physical parts of the types shown and contract a boundary around them, the non-opposed feature (52 width) will allow unlimited contraction - the opposed feature (20 width) will stop contraction at 19.42."

Again, this is because in case of establishment of UAME, "maximum possible contact" or "coincidence with the surface(s) at highest points", although not explicitly stated in the standard, basically means that the feature must be fully immobilized relative to the envelope.

And that is why even though illustration B in the bottom row of my graphic shows the 52 envelope and the actual feature in contact at infinite number of points, this envelope isn't really the UAME of the feature.

Sorry, but that's all I can offer.

### RE: Feature Of Size definition

(OP)

#### Quote (pmarc)

"Take two physical parts of the types shown and contract a boundary around them, the non-opposed feature (52 width) will allow unlimited contraction - the opposed feature (20 width) will stop contraction at 19.42."

pmarc, I don't know if you read my last reply to chez311, but there is a detailed response to the statement you repeated. The chances that you didn't read are high because obviously, you consider me a stubborn nuisance by now. I can only say I did my best to address this, and I would recommend looking at it. If you did read it and chose to ignore it in your last reply it is your right.

The bottom line is that the behavior of UAME simulations for opposed an unopposed features might not be as different as you think. As I've been trying to graphically show and explain in writing, the relative mobility issues between the envelope and the feature are as probable for opposed features as they are for unopposed features. If this isn't true and If they are not equally probable, at least it can be said that there is no guarantee against these issues for both types of features. For these issues to be prevented or minimized, the contraction of the envelope about an external feature or expansion of the envelope within an internal feature has to be deliberately stopped when "maximum possible contact" condition is achieved.

I take the line "Sorry, but that is all I can offer" as a sign that you grow tired of this discourse. Nevertheless, I would appreciate your response because if I'm wrong, I am truly intrigued to understand where specifically my reasoning fails. I will also readily clarify any points I may have not communicated well enough.

### RE: Feature Of Size definition

#### Quote (author unknown, posted 8 Mar 19 12:52 by Kedu)

“E and F are not irregular features of size either. For these two be irregular features of size the mating envelope - two parallel planes - would need to able to close down on these faces. With these geometry closings parallel planes about the part surfaces would pitch ( or rotate) the part and would not actually close down on these faces.”

This seemingly-controversial statement was almost certainly intended to mean that the part would rotate relative to the collapsing envelope, not that the part would rotate relative to some DRF.

I think it's somewhat misleading to say that a part is constrained by a DRF. A DRF is basically just a coordinate system. It would be more accurate to say that the relationship between a part and a DRF is constrained by contact between datum features and datum feature simulators.

Similarly, the relationship between a part and a feature center plane is constrained by contact between the feature and its unrelated actual mating envelope. DRFs are not involved in this.

Moving on to more relevant things:

#### Quote (Sem_D220, 18 Mar 19 22:06)

If you simulate the UAME by a simulator such as a vise which is not attached to anything and is free to translate and rotate in all directions, if you keep tightening the jaws when the separation distance reaches the size of 19.42, the vise will not stay in equilibrium but will keep rotating in the direction which allows further contraction of the distance between the jaws.

That does not appear to me to be the case in your image. The 19.42 envelope appears to contact the top-left, top-right, and bottom-left points. Project these points onto the envelope midplane, and the bottom point ends up between the two top points. This indicates that further contraction is not possible without first expanding and rotating the envelope.

I have created some illustrations showing attempted UAME determination (in 2D) for an external feature nominally consisting of parallel lines 52 units apart:

Image 1 is modeled on the image in the original post, and shows non-opposed surfaces with some waviness but relatively small form error. The plot of envelope width vs. rotation has no local minima (or at least none for alignments anywhere near reasonable), so I conclude that a UAME does not exist for this feature.

Image 2 shows fully opposed surfaces with some waviness but relatively small form error. The plot of envelope width vs. rotation has a single clear valley in which to seek the minimum value, and this minimum value defines the UAME. The depth of the valley and the slope of its walls indicates that the feature is reasonably-well immobilized relative to the envelope.

Image 3 is modeled on the left illustration in the image posted 17 Mar 19 06:07, and shows fully opposed surfaces with relatively large convex form error in addition to some waviness. The plot of envelope width vs. rotation does have local minima, but they are not very prominent. There is no single clear valley in which to seek a minimum value. I conclude that a well-defined UAME does not exist for this feature.

Thoughts?

Also, If anyone would like to take a shot at writing a useful definition for "maximum possible contact", I'd be interested to see that.

pylfrm

### RE: Feature Of Size definition

(OP)
pylfrm,
Note taken regarding the use of terms datum reference frame, datum features and datum feature simulators. It is a good point that what physically constrains the part is datum feature simulators and not the datum reference frame, which is merely a virtual set of 3 perpendicular planes which serves as the reference for measurements and is derivative from the specified datum features in the drawing. I am aware of that and should have probably been more precise in my descriptions.

#### Quote (pylfrm)

This indicates that further contraction is not possible without first expanding and rotating the envelope.

The simulation I performed tells otherwise. I will later try to make clearer figures with enlarged details of the contact zones showing the contraction of the simulator step by step.

The plots you provided look very interesting, but I can't truly relate to them without some visualization of how the movement looked like.
Regardless, image 3 supports the point that the opposed geometry of noncontroversial features of size, doesn't guarantee an unambiguous UAME.

### RE: Feature Of Size definition

(OP)

I hope this illustration will help to visualize how the UAME simulator for a regular opposed as produced FOS might fail if the contraction is not stopped deliberately when maximum contact is achieved. Notice that the UAME envelope will not have to expand and then contracted again at any step to behave as shown. As explained, the contraction is continuous and the rotation occurs as result of the forces applied during the tightening of the simulator about the feature, the simulator simply rotates in the direction which allows it to keep contracting all the way through the process.

As noted previously, the value of 19 is presented as an example. Any envelope size below 19.42 represents a failure to detect the correct UAME size as a result of relying on the physical behavior during contraction about the feature only and trusting on the feature itself to constrain the envelope.

### RE: Feature Of Size definition

#### Quote (Sem_D220)

I take the line "Sorry, but that is all I can offer" as a sign that you grow tired of this discourse.

That line was just my frank admission that I really failed in this thread - I failed because I wasn't able to convince you that you have been wrong.

I read your reply to chez311 (after I submitted my comment), but my only reaction to it can be that in the example with rotating vise your thought process is imply incorrect. As pylfrm explained, once the 19.42 envelope gets in contact with the feature at the top-left, top-right, and bottom-left points, further contraction will be impossible, unless you expand the envelope and rotate it relative to the feature.

To me, the way you have been explaining how further contraction towards 19 and below would be possible is a description of a deliberate action leading to contraction of the envelope about other feature than the feature it should be and could be contracted about. It is like saying that for perfectly manufactured block 20 mm high and for example of 5 mm wide the UAME of the height is 5 mm. This is the path your argument follows, and I truly believe it is a wrong path.

Just in case if you are going to say that the same action is taken in the non-opposed feature example, my answer is that this is not true. In the non-opposed feature case, no matter how hard one tries to precisely contract the envelope about the considered feature, the lack of opposed points will let the envelope to contract further and further. In your example of '19 or less' envelope, one has to first try hard not to contact the considered feature in the intended way to be able to contract the envelope towards 19 and below.

### RE: Feature Of Size definition

pylfrm,

Thank you for your input and the images. I think they objectively (mathematically) explain what chez311 and I have been trying to explain with words throughout the course of this discussion.

The only thing I would add is that the geometry in image 3 is a classic example of a "rocker", and just like in case of convex (nominally planar) datum surface, where a candidate datum concept would be involved (in prior-to-Y14.5-2018 era), the same idea can be applied here. There is a set of candidate unrelated actual mating envelopes, but that doesn't mean the concept of the UAME itself is invalid.

### RE: Feature Of Size definition

(OP)

#### Quote (pmarc)

To me, the way you have been explaining how further contraction towards 19 and below would be possible is a description of a deliberate action leading to contraction of the envelope about other feature than the feature it should be and could be contracted about.

This would not be a deliberate action but the result of relying solely on "physical reality" that suggests that any opposed feature of size will constrain it's actual mating envelope and that the contraction of the envelope will have to be stopped by the feature itself.

Edit: I rechecked the simulstor behavior, no expansion in the process.

### RE: Feature Of Size definition

Feel free to call it as you want. All I am saying is that your process of establishing the '19 or less' envelope is not correct and unavoidably leads to weird conclusions like the one I suggested with the rectangular block 20 mm high X 5 mm wide.

### RE: Feature Of Size definition

(OP)
I never said it was a correct process, and you are right about the weird conclusions that it leads to.
This is what I was trying to say and show all along. Unlike I was told in the beginning of this thread - you can't take "maximum possible contact" out of the UAME concept, not even for regular opposed features of size.

### RE: Feature Of Size definition

#### Quote (Sem_D220)

Unlike I was told in the beginning of this thread - you can't take "maximum possible contact" out of the UAME concept, not even for regular opposed features of size.
Are you just starting over?

### RE: Feature Of Size definition

(OP)
I have no idea what is the reason or the purpose of you asking this and what am I supposed to answer.

Regarding the quote in your response - you stated yourself that 3 points of contact are needed for correct UAME simulation. So you are not even supposed to disagree with this.

### RE: Feature Of Size definition

You didn't have to answer. We apparently disagree with each other and that doesn't seem like going to change. Either it's because we don't understand or don't want to understand each other.

Anyway, I am afraid I will not be able to change it. Thank you.

### RE: Feature Of Size definition

I re-read your statement that I quoted and I agree with it.

### RE: Feature Of Size definition

(OP)
Apparently we don't always disagree. Anyhow, thank you too.

### RE: Feature Of Size definition

#### Quote (Sem D220 18 Mar 19 22:06)

Unless you use excessive force, the object you are inspecting will not translate or rotate in the directions at which the constrained degrees of freedom prevent movement. This is why I brought up the DRF as a subject for consideration after it was stated that for unopposed features, the simulation process will result in movement (rotation) of the part and inability to simulate the UAME as a result of that movement.
I've already stated previously I do not believe the frame of reference matters - part vs. boundary rotating/translating is inconsequential the end result should be the same - the UAME should not be constrained to any DRF. A good gauging setup should be able to handle this (albeit possibly expensive/complex or not feasible for certain cases) - a CMM certainly can.

#### Quote (pmarc 19 Mar 19 11:25)

To me, the way you have been explaining how further contraction towards 19 and below would be possible is a description of a deliberate action leading to contraction of the envelope about other feature than the feature it should be and could be contracted about. It is like saying that for perfectly manufactured block 20 mm high and for example of 5 mm wide the UAME of the height is 5 mm. This is the path your argument follows, and I truly believe it is a wrong path.
I agree with pmarc's statement. See my figure below - depending on the initial orientation of the part in relation to the boundary/simulator during contraction it will either end up at a final boundary of 19.42 on the faces/feature of interest (correct) or of 5 on the opposing faces (wrong). There is no situation where the feature as you have shown it will start in the correct orientation, rotate/contract through angle A all the way to the 19.42 boundary and then continue to rotate/contract to anything less than that. It seems pmarc was unable to convince you of this and if that is the case I doubt my below figure will be enough to do so.

I can however envision a situation where a similar shape would create a situation which you describe. This, I think even more than the convex/barrel shape, will require excessive size/form and possibly orientation error.

### RE: Feature Of Size definition

#### Quote (Sem D220 18 Mar 19 22:06)

As I said several times, the DRF constrains the degrees of freedom of the part/feature, but not of the UAME.

I should clarify that I absolutely have been aware you have been saying this several times. It is the fact that each time you have proceeded to follow it with statements that in my mind, and I think others - though I will not speak for them, contradicts that conclusion which I think bothers me. The difference is that for me the definition ends there - the UAME is not constrained or defined relative to any DRF. Period. No further clarification is necessary - the fact that it is "not constrained to any datum(s)" per 1.3.25.1 means no datums are involved, directly or indirectly.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

I've already stated previously I do not believe the frame of reference matters - part vs. boundary rotating/translating is inconsequential the end result should be the same
I have also said that essentially one can view the rotation of the UAME relative to the part as rotation of the part relative to the UAME. Physically they are the same. But, I always imagine geometrical controls in association with the fixture and inspection process. If the part is constrained in a fixture while the UAME simulated as it would be for the part from fig. 7-65 I posted above, and someone mentions that the UAME (not the RAME, as I hope you now realize) simulator for the slot may move or rotate the part during the interaction (the analogy to that is your sentence "as the boundary closes down the feature will rotate as that boundary closes in" from 8 Mar 19 14:33 ) , I think it is natural to respond by mentioning that the part is constrained in movement by the datum feature simulators A and B (Not by the DRF as I initially described - thank you pylfrm for the correction), and therefore the part can't move. If a ball bounces from the wall I will not say that the wall bounces from the ball. Even though physically speaking the frame of reference doesn't matter in this case too, I say the ball bounces from the wall because the wall is fixed in place and the ball is not. That is the natural way to look at things (at least for me).

If you looked at the figure I posted at 19 Mar 19 10:07 you could see my explanation on how further contraction below 19.42 is possible. I can describe it for the figure you posted: the exact behavior probably depends on the exact as produced angles and there could probably be various scenarios, but if your part dictates a similar UAME simulator behavior, after reaching dimension 19.42, if the "manual" contraction doesn't stop deliberately by the "operator" at exactly that moment, the pivot point around which the top plane of the simulator rotates will translate from the top right corner to the top left corner of the part, while the pivot point around which the bottom plane of the simulator rotates will remain the bottom left corner. This situation allows for a smooth transition and continuous contraction below 19.42.

Edit: quote added in the beginning of the post

### RE: Feature Of Size definition

#### Quote (Sem_D220, 19 Mar 19 10:07)

I hope this illustration will help to visualize how the UAME simulator for a regular opposed as produced FOS might fail if the contraction is not stopped deliberately when maximum contact is achieved. Notice that the UAME envelope will not have to expand and then contracted again at any step to behave as shown.

Correct me if I'm wrong, but I'm going to guess that your example shape is a polygon with vertices at the following coordinates (listed counterclockwise from top right):

p1 = (5, 20 - 5 * tan(7°))
p2 = (0, 20)
p3 = (0, 5 * tan(5°))
p4 = (5, 0)

Between step 2 (19.42 envelope width) and step 3 (19.00 envelope width) in your latest illustration, the envelope must rotate through an orientation where points p1 and p3 are directly opposed. The distance between these points is sqrt(5^2 + (20 - 5 * tan(7°) - 5 * tan(5°))^2) = 19.597212, so the envelope must indeed expand to allow this.

I ran this geometry through my plot-generating program. Image 4 is the result. A discontinuity in the slope is evident at a rotation angle of 5° and an envelope width of 19.49, corresponding to your step 1. Envelope width reaches a local minimum of 19.42 at a rotation angle of 7°, corresponding to your step 2. Getting from there to your step 3 requires going back up and over a local maximum.

What are we missing here?

#### Quote (pmarc, 19 Mar 19 11:25)

The only thing I would add is that the geometry in image 3 is a classic example of a "rocker", and just like in case of convex (nominally planar) datum surface, where a candidate datum concept would be involved (in prior-to-Y14.5-2018 era), the same idea can be applied here. There is a set of candidate unrelated actual mating envelopes, but that doesn't mean the concept of the UAME itself is invalid.

Could you expand on how this candidate UAME concept might be applied, specifically to the geometry in my image 3? I'm particularly interested in what rule might be used to determine what is considered a valid candidate.

pylfrm

### RE: Feature Of Size definition

(OP)
Thank you pylfrm for performing the check. I didn't notice the local maximum that is expected between step 2 and 3. * In case you still have the data saved, could you change the shape to a parallelogram ( two 5° angles instead of a 5° and a 7° ) and check if a local maximum is expected in that case too?

#### Quote (pylfrm)

What are we missing here?

What you (in plural) are missing is that this particular as produced polygon was only intended to serve as an example to communicate a valid point. With additional form deviation at the bottom left corner or a slightly different produced shape the UAME simulator could behave exactly as I showed. In addition, there are the convex "rockers" that don't have a solution, with which pmarc suggests to deal by a method that is intended to be applied on planar features and not features of size.

The point is: the fact that a feature was designed as a simple rectangular feature of size cannot guarantee the ability of the feature to physically restrict the contraction of the envelope simulator about it. And since it can't, the differentiation between features of size and none features of size cannot be based on that. Furthermore, the concept of a contracting envelope until being physically constrained and brought to an equilibrium of forces by the surfaces of the feature is not implied in any form by the definition of the unrelated actual envelope in the standard.

Edit: * never mind about the 5° parallelogram. I have seen that it behaves the same.

### RE: Feature Of Size definition

#### Quote (Sem D220 19 Mar 19 20:58)

But, I always imagine geometrical controls in association with the fixture and inspection process. If the part is constrained in a fixture while the UAME simulated as it would be for the part from fig. 7-65 I posted above, and someone mentions that the UAME (not the RAME, as I hope you now realize) simulator for the slot may move or rotate the part during the interaction (the analogy to that is your sentence "as the boundary closes down the feature will rotate as that boundary closes in" from 8 Mar 19 14:33 ) , I think it is natural to respond by mentioning that the part is constrained in movement by the datum feature simulators A and B (Not by the DRF as I initially described - thank you pylfrm for the correction), and therefore the part can't move.
This may be true in many instances, but not all - it depends on what is being simulated/inspected. There is no requirement that the part ALWAYS be constrained when simulating the UAME. If the desire is to check the UAME against the position tolerance zone then yes the part would have to be fixed relative to the applicable datum features to establish said tolerance zone. If the desire was to check only the size of the UAME for whatever reason there would be no similar requirement and could be perfectly feasible in some instances to have the part unconstrained - ie: for a small, lightweight part it could be held in your hand and a simulator fitted into/onto the applicable feature. I know that for practical purposes you might not want to literally hold it in your hand (thermal expansion from body heat - especially on a small part) but my point remains that there is no requirement to fix it to a specific datum feature(s) in that case and simulation could force the part to move or the part/simulator to move relative to each other.

Most importantly perhaps is the case of a primary datum feature. In that case the part would indeed be unconstrained relative to any other datum features and the simulator (UAME boundary) could be fixed. As the boundary contracts/expands into contact with the primary datum feature the part would be expected to move in relation to the simulator.

It is for these reasons which I stand by my assertion that the frame of reference does not matter - in certain cases the boundary/simulator will move, in others the part, or both. It depends on what is being simulated and/or what is feasible due to part/simulator size/weight/design.

#### Quote (Sem D220 20 Mar 19 04:32)

The point is: the fact that a feature was designed as a simple rectangular feature of size cannot guarantee the ability of the feature to physically restrict the contraction of the envelope simulator about it. And since it can't, the differentiation between features of size and none features of size cannot be based on that.
Perhaps not within every single possible variation or feature configuration (ie: high length/width ratios combined with loose tolerances as in your example) which might require special treatment (fitting routines, etc..) however for the vast majority of cases I would say it does. If what you say was true, standard workholding like a vise would be pretty useless as this is the very simple concept that it works on - contract until it stops. Just because there are a select few cases for which this doesn't work requiring some special treatment is not a reason to stop using a vise for the vast majority of cases where it does work. The same applies to the similar definition of a UAME.

#### Quote (Sem D220 20 Mar 19 04:32)

Furthermore, the concept of a contracting envelope until being physically constrained and brought to an equilibrium of forces by the surfaces of the feature is not implied in any form by the definition of the unrelated actual envelope in the standard.
I am of the opposing opinion - I would say it is the only definition directly supported by the standard. It is an unambiguous interpretation of the statement found in 1.3.25 stating an actual mating envelope is "A similar perfect feature(s) counterpart of smallest size that can be contracted about an external feature(s) or largest size that can be expanded within an internal feature(s)". A minimum/maximum of an envelope occurs when it has contracted/expanded to its fullest extent. Any inclusion of "maximum possible contact" adds no clarity as that term is not defined, in addition to not being in the original definition (your point taken about it being discussed in relation to primary datum references notwithstanding). Any discussion of "equilibrium of forces" is just the physical reality of simulating such a boundary.

### RE: Feature Of Size definition

Additionally, pylfrm's plots show that a minimum can be defined mathematically - no "equilibrium of forces" necessary, which would just be the result of bringing a physical simulator in contact with a physical feature.

pylfrm - thank you for those plots. I may derive a similar program if I have time. I was going to make a series of figures in CAD but your method may prove to be the simpler, as well as more convincing, one.

### RE: Feature Of Size definition

(OP)
chez311,
Regarding the effect of datum feature simulators or lack of them on the interaction between features and UAME envelopes, I agree with the points you brought up. The most important thing is that it remains clear that relative movement problems can be dealt with by technical means and one shouldn't sort features to types according to the probability to experience relative movement problems; for a primary datum feature simulation, the simulator can be fixed, as you said. For position inspection, the part will be fixed during UAME simulation, as you acknowledged. If both the simulator and the part are held manually as in the additional scenario you mentioned, the hands act as the constraining device when needed. Relative movement can and will occur, but it also can and will be controlled.

#### Quote (chez311)

I am of the opposing opinion - I would say it is the only definition directly supported by the standard. It is an unambiguous interpretation of the statement found in 1.3.25 stating an actual mating envelope is "A similar perfect feature(s) counterpart of smallest size that can be contracted about an external feature(s) or largest size that can be expanded within an internal feature(s)".

You ended the quote in the middle of a sentence. The definition is not complete without the missing part: "... so that it coincides with the highest points". This is not just a clarification about the envelope being outside of the material, but part of the requirement: the simulator must surround the feature in a close contact relationship.

This essentially means that if an envelope that conforms to this condition can be found for a given as produced feature, this envelope is the UAME. If the envelope can contract further, but the further contraction is accompanied by loosening the contact with the feature, the envelope that was already established when there was a sufficient contact is still valid, as it remains the smallest envelope that conforms to the definition. There is nothing in the definition that says that it is mandatory that the physical contact between the feature and a simulator constrains the simulator as in a simple case of a vise that is stopped by the object it closes on.

This brings me to pylfrm's plots. It is not always the minimum value that should be looked for. Even where a local minima can't be detected, one of the values can represent the actual UAME size. The plot doesn't provide the information about the amount of contact along the process.

Inspection-wise, it is all a question of the available technology. If detection of the UAME which "coincides with the highest points" (according to pmarc, there should be 3 of them) is possible for a pair of "offset-opposed" surfaces, and nothing in the process contradicts the standard, why forbid it?

### RE: Feature Of Size definition

#### Quote (pylfrm)

Could you expand on how this candidate UAME concept might be applied, specifically to the geometry in my image 3? I'm particularly interested in what rule might be used to determine what is considered a valid candidate.

This is just a quick idea:
https://files.engineering.com/getfile.aspx?folder=...

I am not sure it could be applied to all kinds of rocking actual geometries of this type of regular feature of size, but I think that it could work for the geometry in image 3. Of course, it is not supported by any standard, as far as I can tell, but I would call it a derivative of the procedure of finding a valid candidate datum plane for primary planar datum feature, as defined in Y14.5.1.

### RE: Feature Of Size definition

Sem_D220,

You claim that the underlined portion of the following definition is important:

#### Quote (ASME Y14.5-2009 para. 1.3.25)

envelope, actual mating: this envelope is outside the material. A similar perfect feature(s) counterpart of smallest size that can be contracted about an external feature(s) or largest size that can be expanded within an internal feature(s) so that it coincides with the surface(s) at the highest points.

Can you describe what you think the portion in bold means? Please be as precise as you can.

pmarc,

Thank you for providing the explanation and illustrations.

I think the A = B condition might be problematic, especially for a less symmetrical case such as image 5. I'll have to give this some more thought though.

pylfrm

### RE: Feature Of Size definition

(OP)
pylfrm,
For me, coincidence with the highest points means tangent adjustment between a surface and a plane (edit: or between a surface/feature and an envelope that is of the inverse geometrical form of the surface/feature).

For a planar surface, it is what required when a tangent plane modifier is specified, as depicted in fig. 6-18 in Y14.5-2009.

For a cylindrical feature, fig. 4-11 and 4-12 are probably the best references. Even though they describe datum simulation process, this is essentially a simulation of the UAME.

For a width feature of size consisted of opposed or offset-opposed faces, I think that one of the 2 parallel planes of the simulator should act as a tangent plane to its surface and the other may contact the opposed or offset opposed surface at one point. The smallest (or largest for an internal feature) envelope that can be simulated that way - is the UAME. It is possible that this description is not accurate, or you can find some fault in it. In that case, you can refer to the pmarc's idea that there should be 3 points of contact - 2 on one face and 1 in the other face. The smallest envelope that conforms to this condition is the UAME.

### RE: Feature Of Size definition

#### Quote (Sem D220 21 Mar 19 03:48)

This essentially means that if an envelope that conforms to this condition can be found for a given as produced feature, this envelope is the UAME. If the envelope can contract further, but the further contraction is accompanied by loosening the contact with the feature, the envelope that was already established when there was a sufficient contact is still valid, as it remains the smallest envelope that conforms to the definition.

#### Quote (Sem D220 21 Mar 19 03:48)

For a width feature of size consisted of opposed or offset-opposed faces, I think that one of the 2 parallel planes of the simulator should act as a tangent plane to its surface and the other may contact the opposed or offset opposed surface at one point. The smallest (or largest for an internal feature) envelope that can be simulated that way - is the UAME. It is possible that this description is not accurate, or you can find some fault in it. In that case, you can refer to the pmarc's idea that there should be 3 points of contact - 2 on one face and 1 in the other face. The smallest envelope that conforms to this condition is the UAME.

I have no issue with the concept of requiring three points of contact, I draw issue with it though when that takes precedence "where a local minima can't be detected" as you noted. Even if we were for a moment to take this as the case - the added requirements to what was initially a very simple definition (simulator progresses to its max/min size until it stops) - which I would think already works in the vast majority of cases - to allow a solution in a handful of extreme/minority cases, which could be possibly better served by a different control (ie: profile of offset-opposed planar surfaces) instead of making the definition fit them. Even still, your definition provides no way to handle the convex case, either of the opposed or non-opposed surfaces, which both have the same amount of contact no matter the size of the simulator. I am sure with some effort one could also find other similar cases such as situations where it would be bi-stable with the same boundary size with identical (3 points) contact.

#### Quote (Sem D220 20 Mar 19 20:33)

Inspection-wise, it is all a question of the available technology. If detection of the UAME which "coincides with the highest points" (according to pmarc, there should be 3 of them) is possible for a pair of "offset-opposed" surfaces, and nothing in the process contradicts the standard, why forbid it?
You previously said that you are always thinking of the fixture and inspection process. Consider then a fixture for a primary datum feature of the type in your OP having offset-opposed planar features referenced RMB. Would you say that would make for consistent, reliable primary datum feature able to constrain the required translation/rotation DOF dictated by a centerplane datum? Perhaps in the digital world with a CMM yes but I do not believe it would make a reliable fixture/gauge - especially if there was a case similar to your post 17 Mar 19 10:11 but with offset-opposed planar features, reliably simulating that second minimum with "maximum possible contact" I think would be difficult at best.

#### Quote (Sem D220 20 Mar 19 20:33)

There is nothing in the definition that says that it is mandatory that the physical contact between the feature and a simulator constrains the simulator as in a simple case of a vise that is stopped by the object it closes on.
Actually right there in the section which you referenced about RMB primary datum features (4.11.4) to include a requirement for "maximum possible contact" there is a reference to just this. It states "As a practical example, a machine element that is variable (such as a chuck, mandrel, vise, or centering device) is used to simulate a datum feature simulator of the feature and to establish the simulated datum." While not a "mandatory requirement" I would say it shows pretty clearly that the physical realities of simulation were in the forefront of the committee's mind when this was written and that "the physical contact between the feature and a simulator constrains the simulator as in a simple case of a vise that is stopped by the object it closes on" as you state was exactly what was being considered.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

the added requirements to what was initially a very simple definition (simulator progresses to its max/min size until it stops) - which I would think already works in the vast majority of cases - to allow a solution in a handful of extreme/minority cases, which could be possibly better served by a different control (ie: profile of offset-opposed planar surfaces) instead of making the definition fit them.

The way you put it suggests that I was proposing altering the definition or proposing a definition of my own. Actually, I am basing my interpretation purely on the wording of the standard. As it seems, the concept of *(that the UAME simulation must be based on) "simulator progresses to its max/min size until it stops" is merely a common convention.

An example where position would be a better control than profile is where it is only desired to control a feature of the relevant type for location and orientation, without the form requirement imposed by profile. I did not change the UAME definition to make it fit offset-opposed features.

#### Quote (chez311)

Even still, your definition provides no way to handle the convex case, either of the opposed or non-opposed surfaces, which both have the same amount of contact no matter the size of the simulator. I am sure with some effort one could also find other similar cases such as situations where it would be bi-stable with the same boundary size with identical (3 points) contact.

I did not pretend to offer a solution for any ambiguous cases. I am proposing an interpretation of the feature of size and unrelated actual mating envelope concepts based on the existing definitions in Y14.5-2009. The examples of ambiguous cases were brought with the purpose of communicating the point that the convention that only "classic-opposed" features are features of size doesn't really help to eliminate ambiguity.

#### Quote (chez311)

You previously said that you are always thinking of the fixture and inspection process. Consider then a fixture for a primary datum feature of the type in your OP having offset-opposed planar features referenced RMB. Would you say that would make for consistent, reliable primary datum feature able to constrain the required translation/rotation DOF dictated by a centerplane datum?

It would be up to the designer to decide if an "offset-opposed" feature is reliable enough to be chosen as a primary datum feature. The decision should also be according to the function of the feature in its intended application. The standard doesn't specify that only features that are reliable enough and appropriate to be chosen as primary datum features can be classified as features of size.
Furthermore, I can imagine cases where features of the same type, with larger surface area and perhaps different height to width proportions, can be supported and constrained by mating parts in assembly and can be legitimate datum features.

The practical examples of physical datum feature simulators in para. 4.11.4 are just that - examples. There are numerous references in the standard to digital simulation processes. A quick search for an example provides a portion of para. 4-7:

"In practice, the features are
associated with physical or mathematical elements that simulate the datum feature simulators in a stated order of precedence and according to applicable modifiers."

There is nothing in the standard that suggests that for the establishment of a valid UAME, only capabilities of physical simulation should be considered, or that mathematical methods are less relevant.

* edit: wording in parentheses added.

### RE: Feature Of Size definition

#### Quote (Sem_D220, 21 Mar 19 03:48)

For a planar surface, it is what required when a tangent plane modifier is specified, as depicted in fig. 6-18 in Y14.5-2009.

#### Quote (Sem_D220, 21 Mar 19 03:48)

For a width feature of size consisted of opposed or offset-opposed faces, I think that one of the 2 parallel planes of the simulator should act as a tangent plane to its surface and the other may contact the opposed or offset opposed surface at one point. The smallest (or largest for an internal feature) envelope that can be simulated that way - is the UAME.

Thank you for providing a clear and precise description of your interpretation for width features. I am glad to see that the undefined concept of maximum contact is not involved.

Tangent planes per ASME Y14.5-2009 are handled the same as planar primary datum features, and that subject is covered in detail by ASME Y14.5.1M-1994. As far as I know, the specified procedure always produces at least one candidate datum plane. Therefore I imagine your interpretation would always produce a UAME, although perhaps one I'd consider questionable.

#### Quote (Sem_D220, 21 Mar 19 03:48)

For a cylindrical feature, fig. 4-11 and 4-12 are probably the best references. Even though they describe datum simulation process, this is essentially a simulation of the UAME.

For the these figures, I'd like to point out that contact with highest points is only mentioned in the "Physical datum feature simulator" side of the "Means this" portions. The "Theoretical datum feature simulator" side says that the AME is the smallest circumscribed or largest inscribed cylinder. Similarly, Figs. 4-13 and 4-14 say that the AME for the width feature is the pair of parallel planes at minimum or maximum separation.

pylfrm

### RE: Feature Of Size definition

(OP)
pylfrm,
Thank you for asking the right questions, that helped me to form a better description of my interpretation, using well-defined terms and concepts.

#### Quote (pylfrm)

For these figures, I'd like to point out that contact with highest points is only mentioned in the "Physical datum feature simulator" side of the "Means this" portions. The "Theoretical datum feature simulator" side says that the AME is the smallest circumscribed or largest inscribed cylinder. Similarly, Figs. 4-13 and 4-14 say that the AME for the width feature is the pair of parallel planes at minimum or maximum separation.

I suppose that since the contact instability issues discussed in this thread are mainly expected in a physical simulation process, it makes sense that contact with the highest points is emphasized in the physical simulation side of the figures. If it is agreed that datum simulation for cylindrical and parallel planar surfaces is done by simulating an unrelated actual mating envelope (this appears as a clarification in parentheses in the relevant portions of para. 4.11.4), I think it can be derived that contact with the highest points is also required from the theoretical simulator, as this is part of the general definition of an AME.

### RE: Feature Of Size definition

#### Quote (pylfrm 22 Mar 19 04:11)

Therefore I imagine your interpretation would always produce a UAME, although perhaps one I'd consider questionable.
pylfrm,

Could you expand on what you consider questionable?

In regards to the emphasis on "high points" it seems redundant to me. Going by the simple definition that the simulator/boundary progresses until it reaches a minimum/maximum: if there is a solution (definable local minimum/maximum) or multiple solutions in the case of a "rocker" there will be 3 points of contact with the high points - if there is no solution then contact is irrelevant. It is only when the additional constraint in Sem's interpretation of the min/max separation of the simulator/boundary that ALSO satisfies the 3 points of contact where the distinction becomes important (Sem - if I got your interpretation wrong please correct me, I tried to sum it up succinctly). This, as myself and pmarc have pointed out, creates an unstable boundary as the feature by itself is not able to constrain the boundary as it closes in but instead it must either somehow be artificially constrained against it and contraction/expansion stopped at this point (min/max boundary and 3 points contact).

I don't really have a problem with the idea conceptually - I only have an issue when its presented as inherent in the existing definition as per the wording of the standard. If one has a special case of unopposed-offset planar surfaces I think the concept could be applied as long as it were treated as just that - a special case. Without it and expecting the feature to be interpreted by everyone as the designer expects is asking for trouble. A note or accompanying specification on how it should be treated would I think be more appropriate - if it requires the alternate interpretation as laid out above that should be defined as well as perhaps how it should be affixed/constrained if physical gauging is expected (ie: restrained/clamped somehow to mating unopposed-offset simulators*).

*Actually thinking about this it might not be practical to physically gauge such a feature as it would rely on the inspector to determine somehow (by eye?) when to stop contraction at the exact min/max which also satifies 3 points of contact.

### RE: Feature Of Size definition

(OP)
chez311, please take another look at this paragraph:

#### Quote (ASME Y14.5)

1.3.25 Envelope, Actual Mating
envelope, actual mating: this envelope is outside the material. A similar perfect feature(s) counterpart of smallest size that can be contracted about an external feature(s) or largest size that can be expanded within an internal feature(s) so that it coincides with the surface(s) at the highest points.

You commented on this content:

#### Quote (chez311 13 Mar 19 09:39 )

Perhaps someone else can support/refute this but I believe the reference to the fact that it "coincides with the surface(s) at the highest points" is not a requirement for maximum contact of any sort but instead a clarification/refinement to the requirement that the boundary exist wholly outside the material.

Considering that the first sentence of the definition is the crystal clear statement: "this envelope is outside the material", such "clarification/refinement" at the end of the paragraph seems completely redundant. Doesn't it make more sense as part of the requirement which specifies a certain type of contact/adjustment between the feature and the simulator? Or perhaps you think the committee members were concerned that by the time the reader gets to the end of the paragraph he will forget the beginning of it, so they decided to end the paragraph with a reminder?

#### Quote (chez311)

It is only when the additional constraint in Sem's interpretation of the min/max separation of the simulator/boundary that ALSO satisfies the 3 points of contact where the distinction becomes important ... This, as myself and pmarc have pointed out, creates an unstable boundary as the feature by itself is not able to constrain the boundary as it closes in...

The "additional" requirement of contact at the high points doesn't create unstable boundaries. What may create unstable boundaries is relying solely on the ability of the feature to stop the progression of the simulator. Instead of ruling out features from the feature of size category based on the risk to experience such issues, it may be beneficial to utilize the complete UAME definition of smallest/largest envelope size + contact at the high points.

#### Quote (chez311)

Going by the simple definition that the simulator/boundary progresses until it reaches a minimum/maximum: if there is a solution (definable local minimum/maximum) or multiple solutions in the case of a "rocker" there will be 3 points of contact with the high points

I agree with this part of your statement. Based on this principle I propose the following:
Where it is impossible to achieve minimum/maximum envelope size at which the sufficient amount of contact is maintained by the simple method of contracting/expanding the envelope until it is constrained by the feature, a different method that involves detection of the minimum/maximum envelope size that is still compliant with the "tangent adjustment/ contact at the high points" requirement should be utilized. This approach does not add any complexity to the simple cases but it does add possibilities and is fully compliant with Y14.5-2009. What do you think?

### RE: Feature Of Size definition

Sem,

I hope you don't think I was disregarding your latest response(s). I was just coming up blank with a way to respond without repeating things that had already been said, which is why I appealed to pylfrm to see if another point of view might help and shake things up.

#### Quote (Sem D220 22 Mar 19 19:55)

perhaps you think the committee members were concerned that by the time the reader gets to the end of the paragraph he will forget the beginning of it, so they decided to end the paragraph with a reminder?
Without getting too deep into a word by word grammatical analysis of the paragraph, I don't think they are fully redundant - overlapping yes, but not redundant. If the first statement about the boundary being outside the material is removed, there is no reference for what constitutes the highest points. If the second statement about contact with the highest points is removed, there is no other requirement that the boundary make contact with the feature at all - other than that implied by the envelope be contracted/expanded to its smallest/largest size. That may be nitpicking a bit, but being that its a standard which should strive to be as precise as possible and leave little to implication with important definitions (whether or not it succeeds in that respect in every case is another matter). I guess the immediate response would be why clarify highest points in that case and not just leave it at "coincides with the surface" as a boundary which exists wholly outside the material can only possibly contact a surface at the "highest points" - for this I don't have a good answer, I can only chalk it up to the committee trying to be as precise as possible. I just don't see it as an additional requirement for amount of contact which would take precedence in the case where no local minimum could be found.

I know that I said previously the emphasis on the high points was redundant, which might serve to discount my statements above. Perhaps I had a poor choice of words. What I meant was that contact with the high points seemed to me to inherently follow directly from determination of a local minimum/maximum, not instead as an additional requirement.

#### Quote (Sem D220 22 Mar 19 19:55)

The "additional" requirement of contact at the high points doesn't create unstable boundaries. What may create unstable boundaries is relying solely on the ability of the feature to stop the progression of the simulator.
I say it creates an unstable boundary because progression of the simulator must be stopped in certain cases when it is determined that the 3/high point contact and minimum/maximum separation are both satisfied, which means that the feature does not fully restrain the boundary. Considering a physical example would you agree that in the case of a feature having 2x unopposed-offset planar surfaces would not fully restrain the simulator which it contacts? There is nothing physically preventing further progression of the envelope or rotation of the feature away from the simulator (or the simulator away from the feature - lets not get into frame of reference again..). Relying solely on the ability of the feature to stop progression of the simulator does not create unstable boundaries - it rejects them (ie: if there is no local minimum/maximum as in the unopposed-offset case then no UAME exists). I have already conceeded that in certain cases where nominally what is typically a "well-behaved" feature may create issues when large variation is allowed, special treatment/consideration may be required.

Note also that there are several examples of irregular FOS in the standard similar to what we have been discussing. In Y14.5-2009 this would be 4-33/4-34/4-35 and in Y14.5-2018 these same figures correspond to 7-40/7-41/7-42, the first two of which are now directly referenced in the definition for irregular FOS for 2018. These all consist of 3x unopposed-offset features which are inherently stable and only require simple contraction/expansion of the simulator to its minimum/maximum limit, instead of 2x in your case - I do not think this is an accident.

#### Quote (Sem D220 22 Mar 19 19:55)

Where it is impossible to achieve minimum/maximum envelope size at which the sufficient amount of contact is maintained by the simple method of contracting/expanding the envelope until it is constrained by the feature, a different method that involves detection of the minimum/maximum envelope size that is still compliant with the "tangent adjustment/ contact at the high points" requirement should be utilized. This approach does not add any complexity to the simple cases but it does add possibilities and is fully compliant with Y14.5-2009. What do you think?
I've already said that I don't really have an issue with imparting this requirement in special cases, through a note or supplementary specification. I don't think it contradicts anything in the standard (by that I mean provides a conflicting definition), just an additional requirement that is not originally there. The only issue I have is stating that this is the default.

### RE: Feature Of Size definition

#### Quote (chez311, 22 Mar 19 13:41)

Could you expand on what you consider questionable?

For a feature similar to image 2 (fully opposed, no extreme form error), it's possible that neither boundary of the minimum-width envelope would be a valid tangent plane for its corresponding surface. Adding that requirement would force selection of a larger-width envelope (which may not even be a local minimum) for the UAME. This strikes me as undesirable.

Also, I don't see much value in a UAME being defined at all for non-opposed features.

pylfrm

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

I just don't see it as an additional requirement for amount of contact which would take precedence in the case where no local minimum could be found.

I have always understood the expression "contact on the highest points" as a description of a specific geometrical relationship between a simulated boundary and the feature. I have seen this wording utilized in other sources, not just the standard, and I think the committee members use this wording to specify a concrete condition. Look at the text under fig. 6-18, that describes the tangent plane concept. If the illustration would be removed and we had only the text left, the exact type of contact illustrated there would have to be understood from the wording "A plane contacting the high points of the surface".

Another reason why I doubt that the concept of contraction/expansion of the simulator until physically being brought to mechanical equilibrium by the surfaces of the feature, or alternatively the method of looking for a local minima, is the essence of the UAME definition, is that it would mean that a concept in Y14.5 is defined based on a gauging technique or a data analysis method. As you know, Y14.5 doesn't deal with physical gauging or mathematical definitions, and generally, it doesn't base any concepts or requirements on these. Where references to gauging equipment are provided in Y14.5, it is only as examples for explanatory purposes. That is why the operation principle of a vise can't be the essence of the AME definition. It can probably be brought up as an implementation example that can cover the majority of cases, but it is not the definition itself.

### RE: Feature Of Size definition

(OP)
pylfrm, I may be missing something but I can only imagine that a stable minimum envelope for a width feature will always consist of a tangent plane contacting at least 3 high points and a plane parallel to the tangent plane contacting at least one high point. The other scenario would be the case of slightly convex surfaces where only one point of contact (edit: at each side) may be maintained for various possible unstable minimum envelopes.

### RE: Feature Of Size definition

#### Quote (Sem D220 27 Mar 19 04:53)

that it would mean that a concept in Y14.5 is defined based on a gauging technique or a data analysis method. As you know, Y14.5 doesn't deal with physical gauging or mathematical definitions, and generally, it doesn't base any concepts or requirements on these.

Theres a fine line between specifying a gauging/measurement technique and developing definitions which take into account the physical realities of assembly and inspection. I would say that Y14.5 generally does NOT do the former (with some exceptions) and tries to accomplish the latter (again - with some exceptions). Just substitute for vise a more general definition of "any physical gauge consisting of parallel planar/flat simulators which contract/expand upon a feature".

Also generally you would be correct that Y14.5 avoids specifying gauging/measurement techniques - but lets not forget the notable exception that runout is based on, and actually specifies*, use of an indicator on a rotated workpiece. Now obviously this can be simulated in other ways (ie: on a CMM) but it does not detract from the fact that a measurement/gauging technique drove the definition of a type of tolerance and is specified* in the body of the standard.

*Edit: Please don't take my use of the word "specify" too literally. I understand the standard doesn't actually require use of an indicator, and says "It is neither the intent nor within the scope of this Standard to define measurement methods".

### RE: Feature Of Size definition

pylfrm,

Is something like the below what you had in mind? The bottom surface has constant curvature while the top surface is curved in the middle between points A and B, and has two inclined planar surfaces on either side of A and B. 2 points of contact is achieved at the minimum separation where with 3 points of contact you no longer have a minimum and the UAME is at a perhaps less than optimal or expected orientation.

Edit: realized I referenced the wrong letters for the points

### RE: Feature Of Size definition

#### Quote (Sem_D220, 27 Mar 19 05:49)

I may be missing something but I can only imagine that a stable minimum envelope for a width feature will always consist of a tangent plane contacting at least 3 high points and a plane parallel to the tangent plane contacting at least one high point.

I am going to ignore the word "tangent" for a moment and concentrate on the "stable minimum envelope" and contact point subject.

For the 2-dimensional case, I will define the following:

Condition 1: One surface contacts the envelope at two points (call these a1 and a2) and the other surface contacts the envelope at one point (call this b1).

Points a1p, a2p, and b1p are the orthogonal projections of points a1, a2, and b1 onto the envelope midline.

Condition 2: Point b1p is between points a1p and a2p.

For the 3-dimensional case, I will define the following:

Condition 1 (option A): One surface contacts the envelope at three points (call these a1, a2, and a3) and the other surface contacts the envelope at one point (call this b1).

Condition 1 (option B): One surface contacts the envelope at two points (call these c1 and c2) and the other surface also contacts the envelope at two points (call these d1 and d2).

Points a1p, a2p, a3p, and b1p (option A) or c1p, c2p, d1p, and d2p (option B) are the orthogonal projections of points a1, a2, a3, and b1 (option A) or c1, c2, d1, and d2 (option B) onto the envelope midline

Condition 2 (option A): Point b1p is within the triangle having points a1p, a2p, and a3p as vertices.

Condition 2 (option B): The line segment connecting points c1 and c2 crosses the line segment connecting points d1 and d2.

I'd say a stable minimum envelope is one that corresponds to a local minimum on the width vs. angle plots I've been posting. Such an envelope will always satisfy conditions 1 and 2. I think it's more useful to look at these conditions as consequences instead of requirements though.

#### Quote (chez311, 27 Mar 19 15:22)

Is something like the below what you had in mind?

In you example, both boundaries of the minimum-width envelope are valid tangent lines. See ASME Y14.5.1M-1994 para. 4.3.2 regarding candidate datum sets for nominally flat datum features, although that definition must be converted to 2D for the examples we're looking at here.

I had in mind a feature such as the following:

   surface 1:
point 1:  (0.00, 0.40)
point 2:  (1.00, 0.40)
point 3:  (2.00, 0.37)
point 4:  (3.00, 0.35)
point 5:  (4.00, 0.34)

surface 2:
point 6:  (0.00, -0.40)
point 7:  (1.00, -0.40)
point 8:  (2.00, -0.37)
point 9:  (3.00, -0.35)
point 10: (4.00, -0.34) 

The minimum envelope width is 0.80, obtained with contact at points 1, 2, 6, and 7. For each surface, all contact points are within (1/3)*(4.00-0.00) of one end. This means that neither boundary is a valid tangent plane per the ASME Y14.5.1M-1994 definition.

An envelope width of 0.819836 can be obtained with contact at points 2, 5, and 6 (or 1, 7, and 10), and the boundary that contacts two points is a valid tangent plane. This is the smallest envelope that meets the additional requirement, but it not stable in the sense described above.

pylfrm

### RE: Feature Of Size definition

(OP)
chez311,
In your latest sketch, boundary 20 would not be stable. It would rock around R40 and R62.535 similarly to the "pure" convex feature scenario we were discussing. Boundary 20.026 would have more points of contact with the faces and therefore more stable, and per my interpretation more in line with the specification "so that it coincides with the surface(s) at the highest points."

pylfrm, thank you for the detailed example.

#### Quote (pylfrm)

The minimum envelope width is 0.80, obtained with contact at points 1, 2, 6, and 7. For each surface, all contact points are within (1/3)*(4.00-0.00) of one end. This means that neither boundary is a valid tangent plane per the ASME Y14.5.1M-1994 definition.

Would you say that validation according to the candidate datum set method is required for every single case where a tangent plane determination is required? If so, why?

### RE: Feature Of Size definition

I should clarify that the statements about contact points in my previous post are only intended to be valid when the actual part surfaces are seen as finite collections of points, not continuous smooth curves or surfaces.

Also, "onto the envelope midline" should actually be "onto the envelope midplane" for the 3-dimensional case.

#### Quote (Sem_D220 28 Mar 19 08:24)

chez311,
In your latest sketch, boundary 20 would not be stable. It would rock around R40 and R62.535 similarly to the "pure" convex feature scenario we were discussing.

I disagree. That is the unique envelope of minimum width, so rocking would not be possible without an expansion of the envelope. Earlier examples behaved differently due to greater curvature of the surfaces. This was illustrated and discussed in the 18 Mar 19 16:54 post by chez311.

#### Quote (Sem_D220 28 Mar 19 08:24)

Boundary 20.026 would have more points of contact with the faces and therefore more stable, and per my interpretation more in line with the specification "so that it coincides with the surface(s) at the highest points."

Contact points between actual surfaces in the real world cannot be counted. Contact points between an actual surface and a theoretical envelope can perhaps be imagined, but I don't see how a definition that involves counting them would be useful.

#### Quote (Sem_D220 28 Mar 19 08:24)

Would you say that validation according to the candidate datum set method is required for every single case where a tangent plane determination is required? If so, why?

Methods are not within the scope of these standards. Validate the tangent plane with whatever method is appropriate for the task at hand. That doesn't change the definition though.

There's a significant difference between identifying one valid tangent plane and identifying all valid tangent planes. The former is often much easier, but it's not sufficient for determination of the UAME with your proposed interpretation.

pylfrm

### RE: Feature Of Size definition

(OP)

#### Quote (pylfrm)

Contact points between actual surfaces in the real world cannot be counted. Contact points between an actual surface and a theoretical envelope can perhaps be imagined, but I don't see how a definition that involves counting them would be useful.

Paragraph 4.10.1 "Development of a Datum Reference Frame for Parts With Planar Surface Datum Features" defines the minimum number of contact points required between each planar datum feature and its datum feature simulator according to the datum precedence order. If the gauging is done physically, I agree that counting the points of contact is not practical and not needed. It is assumed that the defined minimum number of contact points will be achieved mechanically by stabilizing the part on the datum feature simulator surfaces according to the sequence prescribed by the datum precedence order. If the simulation is done with CMM I assume that the number of contact points between a simulated plane and the scanned surface extremities is a parameter in the simulation process, and that is when a definition that mentions the required number becomes useful.

#### Quote (pylfrm)

Methods are not within the scope of these standards. Validate the tangent plane with whatever method is appropriate for the task at hand. That doesn't change the definition though.

There's a significant difference between identifying one valid tangent plane and identifying all valid tangent planes. The former is often much easier, but it's not sufficient for determination of the UAME with your proposed interpretation.

Perhaps I should have asked my question more directly: For your example of the feature with the specified coordinates, you said that the 0.8 envelope consists of planes which are not valid tangent planes according to the ASME Y14.5.1M-1994 definition because all contact points at each side lie within 1/3 of the surface from the end. I am barely familiar with ASME Y14.5.1M, but recognize the requirement from the projections of all contact points on a line along the simulated plane not to be within 1/3 of the length from the end of that line, as part of the candidate datum set concept. If it is what you meant, why do you apply the candidate datum set requirement on a tangent plane where this tangent plane is not intended to be used as a datum? Would also you require validation of the simulated tangent plane according to candidate datum set for the example shown in fig. 6-18? In the context of UAME simulation for a width feature, considering my interpretation that one of the simulated planes should be a tangent plane, I don't see a reason to treat it similarly to a datum plane.

### RE: Feature Of Size definition

#### Quote (Sem D220 29 Mar 19 12:13)

Paragraph 4.10.1 "Development of a Datum Reference Frame for Parts With Planar Surface Datum Features" defines the minimum number of contact points required between each planar datum feature and its datum feature simulator according to the datum precedence order.

#### Quote (Sem D220 29 Mar 19 12:13)

why do you apply the candidate datum set requirement on a tangent plane where this tangent plane is not intended to be used as a datum? Would also you require validation of the simulated tangent plane according to candidate datum set for the example shown in fig. 6-18? In the context of UAME simulation for a width feature, considering my interpretation that one of the simulated planes should be a tangent plane, I don't see a reason to treat it similarly to a datum plane.

While it refers to a minimum number of points - the above referenced para 4.10.1 as well as the section about tangent planes (para 6.5) which you referred to in your definition of contact with the high points both refer back to Y14.5.1 for rocking/convex surfaces. The minimum 3 points of contact is the ideal case, but as pylfrm pointed out about my example a convex feature (designed to be flat) can have less than those 3 points of contact and still derive a (or set of) canditate datum plane(s) as well as tangent plane(s). To deal with rocking/convex surfaces in Y14.5.1 there are the sections 4.3.2 (planar) that pylfrm noted, 4.3.3, 4.3.4, 4.3.5 (FOS - RFS/MMC/LMC respectively) - these all deal with datum features and there is no other section as far as I can tell which deals with rocking/convex surfaces. Pylfrm can correct me if I'm wrong, but I think this is why the candidate datum concept is being applied here. It makes sense that this would apply to features which are not datum features as well - if a primary FOS datum is referenced the UAME is the envelope of interest - this should be the same UAME (or set of valid envelopes) if the feature was not referenced as a datum. Note that in several of your responses going as far back as your post on 14 Mar 19 03:46 it seems you have utilized similar logic in relating definition of the UAME to definition of a primary RFS datum - clearly the connection is not lost on you either.

Additionally I don't see anything in this definition of candidate datums about number of points of contact defining a more or less valid candidate datum.

### RE: Feature Of Size definition

#### Quote (pylfrm 28 Mar 19 01:57)

In you example, both boundaries of the minimum-width envelope are valid tangent lines. See ASME Y14.5.1M-1994 para. 4.3.2 regarding candidate datum sets for nominally flat datum features,
Thank you pylfrm for pointing this out - I did not make the connection between tangent planes and candidate datum sets until I read this. I was initially responding to Sem's assertion about the number of points of contact, but since your post I'm even less convinced of the need as you pointed out to "count" the number of points of contact.

### RE: Feature Of Size definition

#### Quote (Sem_D220, 29 Mar 19 12:13)

If the simulation is done with CMM I assume that the number of contact points between a simulated plane and the scanned surface extremities is a parameter in the simulation process, and that is when a definition that mentions the required number becomes useful.

Consider a different case: determination of the maximum inscribed sphere for a set of points. There will always be four contact points, but that's just a consequence of the geometry. It would not make sense for the definition of the UAME of an internal spherical surface to explicitly require four contact points.

I'd apply the same argument to external width features. A definition based on minimization of envelope width seems much more meaningful than one based on counting contact points.

#### Quote (Sem_D220, 29 Mar 19 12:13)

If it is what you meant, why do you apply the candidate datum set requirement on a tangent plane where this tangent plane is not intended to be used as a datum? Would also you require validation of the simulated tangent plane according to candidate datum set for the example shown in fig. 6-18?

I think the candidate datum set definition would apply in all cases where a tangent plane is required (including Fig. 6-18) because para. 2.16 (tangent planes) refers to para. 4.11.2 (datum features) which refers to the Y14.5.1M definition.

What do you think the UAME should be for the feature with coordinates posted 28 Mar 19 01:57?

pylfrm

### RE: Feature Of Size definition

(OP)

#### Quote (pylfrm)

I think the candidate datum set definition would apply in all cases where a tangent plane is required (including Fig. 6-18) because para. 2.16 (tangent planes) refers to para. 4.11.2 (datum features) which refers to the Y14.5.1M definition.

What do you think the UAME should be for the feature with coordinates posted 28 Mar 19 01:57?

Paragraph 2.16 says: "If the tangent plane is unstable it may be optimized. See para. 4.11.2 and ASME Y14.5."
It doesn't require optimization in every single case. So unless there is an instability situation for the 0.80 envelope for the feature with coordinates posted 28 Mar 19 01:57, I wouldn't require candidate datum set validation, and say that envelope 0.80 is the UAME envelope.

#### Quote (pylfrm)

Consider a different case: determination of the maximum inscribed sphere for a set of points. There will always be four contact points, but that's just a consequence of the geometry. It would not make sense for the definition of the UAME of an internal spherical surface to explicitly require four contact points

I'm not sure about that. Let's start with a simpler case of a 2D maximum inscribed circle for a given set of points in 2D. Unless there are 3 points located on a perfect radius, which is not likely, the maximum inscribed circle will have to pass through 2 points. If it passes only through 1 point, it is not really the maximum inscribed circle. So here we see that the number of "contact" points is meaningful.

### RE: Feature Of Size definition

(OP)

#### Quote (pylfrm 29 Mar 19 00:23 )

I disagree. That is the unique envelope of minimum width, so rocking would not be possible without an expansion of the envelope. Earlier examples behaved differently due to greater curvature of the surfaces.

This is another point I wanted to address and I finally got to it now - according to my check, rocking of +-1° of the envelope planes around the two radii on the opposed faces would produce an increase of the envelope by as little as 0.013mm (from 20 to 20.013). So from a purely theoretical aspect, your assertion is correct - it is impossible for the envelope to rock without enlarging, but practically I doubt the ability to determine an envelope that behaves this way repeatably.

If we were to simulate a datum center plane from this feature, if I may return for a minute to the notorious requirement in para. 4.11.4 for the undefined "maximum possible contact", then despite the lack of a definition, it is obvious that the envelope of size 20.026 makes more contact with the feature than the 20 envelope. So, which would be the more valid UAME envelope for a primary datum center plane simulation?
Should we ignore a requirement specified in the standard just because it uses a term not defined well enough?

Edit: the subject of this post is the figure posted by chez311 at 27 Mar 19 15:22.

### RE: Feature Of Size definition

#### Quote (Sem_D220, 1 Apr 19 11:22)

it is obvious that the envelope of size 20.026 makes more contact with the feature than the 20 envelope.

Is it really?

Imagine we measure a bunch of points on this feature with a CMM (with some small random measurement error), and then find the minimum-width envelope of those points (in 2D). Perhaps 19.999 is the result. This envelope will contact two measured points from one surface, and one measured point from the other surface.

Now imagine we rotate the envelope through a range of approximately +/- 5 degrees from the minimum width. We'll have a close approximation to the theoretical 20.026 envelope at somewhere around 1.43 degrees of rotation. Over the entire 10 degree range, the envelope will almost always contact only two measured points. Occasionally it will contact three points, but not more.

pylfrm

### RE: Feature Of Size definition

Sem,

See my figure below. While I also think that a convex surface like this with small amount of form error will have issues with repeatably physically simulating that minimum 20 boundary, I do not think that the additional requirement for contact which you have stipulated (resulting in the 20.026 boundary) imparts any increase in stability. So long as we are discussing the physical realities of inspection/simulation - as your boundary rotates, in addition to increasing in size as pylfrm mentioned (however infinitesimally), the line which is normal to the planar boundary on each side (this envelope being tangent to the convex feature surface) increases in separation from the contact points of the boundary on the opposing side. This is represented in my figure by line D, intersecting with the convex surface (of R62.525) and the planar boundary at point C, which intersects with the opposing boundary at point E. Since the closest contact point on the opposing surface is at point A, I would think that the part/boundary would want to pivot around this point towards the minimum envelope as the boundary contracts since the physical reality would be that the simulator would exert a force on the surface creating a moment about this point.

That being said, I do agree that there would be issues with repeatably simulating the theoretical minimum of 20, as there might be with any convex feature - hence the need for a stabilization procedure through the candidate datum concept. In reality I think I would expect the result to be somewhere between 20.026 and 20, the point being though that the "increased" contact at 20.026 does not in my opinion result in any more stable an envelope (or make it any easier for the operator to find the elusive point of "maximum contact") and during simulation I think you would find that the resulting simulated envelope would be biased toward the minimum 20 envelope.

Finally, I rechecked the figure after pylfrm's notes about candidate datums/tangent planes and it just so happens that the points A and B where the planar portions of the upper surface meets the convex portion lie within 1/3 of the width from either end of the feature. Unless I am missing some nuance of the concept, I do not believe the boundary on either side of the 20.026 envelope would be valid tangent planes.

### RE: Feature Of Size definition

(OP)
pylfrm, If we are talking 2D,
The 20 (or 19.999) envelope should contact one point at one side and one point at the opposite side - those are the apexes of the two radii in chez311's figure.

At 1.43° of rotation, as you say, is the approximate 20.026 envelope where theoretically the top plane of the simulator lies flat on the flat section of the top face. In practice, this is equivalent to 2 contact points at 2D.

In 3D, I don't think it would be hard to find an orientation where the top plane contacts the flat section of the top face at 3 points. Then the other plane parallel to it will contact the bottom arc-shaped face at 1 point. For the approximate 20 envelope, this amount of contact is not guaranteed.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

Since the closest contact point on the opposing surface is at point A, I would think that the part/boundary would want to pivot around this point towards the minimum envelope as the boundary contracts since the physical reality would be that the simulator would exert a force on the surface creating a moment about this point.

chez311, that is true for physical gauging simulation, and for physical gauging, both envelopes would have stability issues (the 20 as you agreed, and the 20.026 as you just rightfully pointed out).
If the simulation is done with CMM and an appropriate computer program, there's a substantial possibility to find an orientation that conforms to the "maximum possible contact"/"contact on the high points"/"tangent plane" (choose the least-worse wording) condition at one of the 2 orientations corresponding to the approximate 20.026 envelope, because this is where a plane contacts a more or less flat portion of a surface. This still doesn't lead to a single solution as there are going to be 2 of them, but it doesn't ignore the required amount of contact implied by the definitions in the standard.

In addition, I am not sure that the candidate datum set concept must be applied in this case. It probably must be for physical simulation - with that I agree.

### RE: Feature Of Size definition

Sem,

I don't think I ever said the minimum 20 envelope had stability issues - I said I agreed it may be difficult to simulate repeatably. I would say it is the most stable envelope that could be derived from the feature as shown.

#### Quote (Sem D220 2 Apr 19 16:31)

In addition, I am not sure that the candidate datum set concept must be applied in this case

Thinking about it, it seems to me that the candidate datum set concept is probably always in effect. Its just whether that datum set consists of no solution, a single stable solution, or multiple solutions. I don't think you can have a solution/tangent plane/candidate datum plane which is valid for a given feature/surface but not a valid candidate datum.

### RE: Feature Of Size definition

(OP)

#### Quote (chez311)

I don't think I ever said the minimum 20 envelope had stability issues - I said I agreed it may be difficult to simulate repeatably. I would say it is the most stable envelope that could be derived from the feature as shown.

My bad. It doesn't necessarily have stability issues, but it does have repeatability issues, as you agreed. I acknowledge the difference, even though often repeatability of measurements/simulations is tightly related to stability. Whether or not a case of some additional form error (probably inevitable in reality) will cause the considered feature to become unstable in interaction with the size 20 envelope - probably better not to get into it.

#### Quote (chez311)

Thinking about it, it seems to me that the candidate datum set concept is probably always in effect. Its just whether that datum set consists of no solution, a single stable solution, or multiple solutions. I don't think you can have a solution/tangent plane/candidate datum plane which is valid for a given feature/surface but not a valid candidate datum.

Y14.5 only refers us to candidate datum set where rocking issues are encountered. More specifically where the as produced geometry of a planar datum feature or a surface for which a tangent plane is required cause the instability of a physical datum feature simulator. What prescribes the application of the candidate datum set for the following cases?
1. Computerized datum/tangent plane simulation for a scanned feature where physical stability does not come into play.
2. Even more importantly - physically stable envelopes where all projected contact points fall within the near-edge 1/3 of the length of a line along the candidate datum plane, such as in the case of the feature described with coordinates by pylfrm at 28 Mar 19 01:57.

### RE: Feature Of Size definition

#### Quote (Sem D220 2 Apr 19 20:07)

What prescribes the application of the candidate datum set for the following cases?

Y14.5.1 governs the mathematical definition of many of the concepts in Y14.5 and is the basis for CMM measurement/software. The fact that it is only referenced 2x in the body of Y14.5 in conjunction with rocking of datum features and 1x in conjunction with circularity does not mean it should only be applied in those cases. I would say as I said before, the candidate datum set contains every solution of candidate datums for a given feature/surface, not just ones which qualify as "unstable" or "rocking" - this can be either 0,1,or greater than 1. This is actually alluded to in Y14.5.1 and "rocking" is only mentioned as a case where there is more than one solution, even a perfectly flat surface* has a candidate datum set - it just so happens that this set contains a singular solution.

#### Quote (Y14.5.1-1994)

4.3.2 Planar Datum Features. The candidate datum set for a nominally flat datum feature is defined in a procedural manner. This empirical defintion specifies a set of datums which are reasonable from a functional standpoint. If the datum feature is perfectly flat, the candidate datum set consists of only one datum; otherwise it may consist of more than one datum. This is equivalent to “rocking” the datum feature on a perfect surface plate

*Edit - missed a word

### RE: Feature Of Size definition

#### Quote (Sem_D220, 2 Apr 19 15:42)

If we are talking 2D,
The 20 (or 19.999) envelope should contact one point at one side and one point at the opposite side - those are the apexes of the two radii in chez311's figure.

This would be true of smooth curves, but not sets of points like I described. If you have an envelope with one contact point on each boundary and those points are directly opposed, then rotation in either direction will allow the envelope width to decrease. If the two contact points are slightly offset, then rotation in whichever direction increases that offset will allow the envelope width to decrease further.

#### Quote (Sem_D220, 2 Apr 19 15:42)

At 1.43° of rotation, as you say, is the approximate 20.026 envelope where theoretically the top plane of the simulator lies flat on the flat section of the top face. In practice, this is equivalent to 2 contact points at 2D.

Yes, just like we had with the 19.999 envelope. The points might be farther apart, but the count is the same.

Essentially the same arguments apply in 3D. The minimum-width envelope will have four total contact points (3+1 or 2+2), and an envelope oriented differently won't have any more than that.

pylfrm

### RE: Feature Of Size definition

(OP)
pylfrm, intuitively it seems that the more or less flat portion of the surface is more likely to allow tangent plane or "maximum possible contact" simulation, but perhaps you are right. Anyway, I agree with you and chez311 that the definition in Y14.5 for the required amount of contact between the feature and the simulator during UAME simulation is missing, or vague at best.

I suppose that the foregone conclusion from this discussion is that offset-opposed surfaces are better not to be treated as features of size.

Thank you chez311, pylfrm, pmarc, Kedu, axym and the others for the great input.

#### Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

#### Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

#### Resources

eBook - Efficient and Effective Production Support with 3D Printed Jigs and Fixtures
Jigs and fixtures offer manufacturers a reliable process for delivering accurate, high-quality outcomes, whether for a specific part or feature, or for consistency across multiples of parts. Although the methodologies and materials for producing jigs and fixtures have evolved beyond the conventional metal tooling of years past, their position as a manufacturing staple remains constant due to the benefits they offer. Download Now
Overcoming Cutting Tool Challenges in Aerospace Machining
Aerospace manufacturing has always been on the cutting edge, from materials to production techniques. However, these two aspects of aerospace machining can conflict, as manufacturers strive to maintain machining efficiency with new materials by using new methods and cutting tools. Download Now

Close Box

# Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

• Talk To Other Members
• Notification Of Responses To Questions
• Favorite Forums One Click Access
• Keyword Search Of All Posts, And More...

Register now while it's still free!