Orientation of a center plane of a tapered feature 2

[Burunduk]

A designer applied perpendicularity feature control frame with the leader pointing to a centerline representing the center plane of an internal taper feature (a "pocket" with symmetrical non-parallel opposed surfaces). I don't consider this specification valid according to ASME Y14.5-2009 but I'm struggling to provide a good explanation of why perpendicularity shouldn't be used this way. The one use of perpendicularity I know when it is applied on a virtual, derived geometry (as opposed to an actual surface) is when a center plane/axis of a feature of size is controlled. It doesn't seem right in the context of a tapered feature, not associated with a size dimension - but I can't form a good argument why. A valid point is that a center plane can be derived from a tapered feature (for example, a datum plane derived from tapered datum feature), and I suppose that a way to evaluate the derived plane orientation relative to a DRF can be found. I need your help, please.

 

[dtmbiz]

Chez311,

It’s been awhile since I reviewed this thread.

Regarding the reference to “being confused”, that may have been directed toward you however Burunduk has since corrected me. In any event the comment wasn’t meant to be “harmful”. It was meant to be a reminder to myself as well.

"Infinity", "the standard is incorrect", "improperly defined", "unfortunately uses" are all useless in advancing the application of dimensioning and tolerancing to finite features and
parts. ASME Y14.5 2009 is the governing standard, its not perfect however it is what we have.

"Useless" is used in reference to the expansion of Y14.5's range, with far reaching terms. I would never want to dishearten or baffle someone because of their personal intellect or reasoning.
The statement was made in an attempt to rein in the discussion for alignment with the terms defined in Y14.5. Now that was "useless".
As a contributor to the thread, "to me" there seems to be a great deal of thought (usually) in your posts which I do not dismiss.

If a Y14.5 discussion is not constrained and connected to the accepted document's principles, terms and their definitions, rules, etc.
then that discussion has little chance for advancement of understanding Y14.5 or arriving at agreement regarding the document (Y14.5)

"To me", after a quick overview of the thread, it seems to a have put significant discussion on "behavior" vs. "definition of terms" which complicates discussion. "To me" this again complicates discussion.
Building blocks would be useful. First understanding the definition and it's intended use and then "behavior" which I take is in reference to an action as how to arrive at the definition.

For instance, "True Profile", "boundary", and "envelope" have specific meanings within the context of Y14.5. They are not interchangeable. Discussion is complicated even more when they are interchanged.

"To me" it is similar to legal terminology in the sense that lawyers in a court room are required to write briefs and address a judge with accepted legal terms.
"To me' Y14.5 is similar in that when in a Y14.5 "forum" for discussion accepted terms especially for the significance to the terms meanings are extremely important.

We all do slip at times. I can remember when in one of my Y14.5 classes how hard the instructor would come down on students when datum and datum feature were misused in their questions.
Seemed rough at the time but it was for our own good to emphasize and reinforce Y14.5 terms definitions and the significant differences in their meanings.

 

[chez311]

As a contributor to the thread, "to me" there seems to be a great deal of thought (usually) in your posts which I do not dismiss.

Well thats a backhanded compliment if I've ever seen one. Thanks?

"True Profile", "boundary", and "envelope" have specific meanings within the context of Y14.5. They are not interchangeable.

Despite your continual insistence that there is a concrete and exacting definition for such terms only one of those three is explicitly defined in the standard - this is "true profile". I've already repented for and retracted my statement several times referring to the true profile as a boundary.

The second two are only defined contextually with other concepts ie: Maximum Material Boundary or Unrelated Actual Mating Envelope. If you believe the words boundary and envelope have individual, special, or sacred meaning you should probably alert the ASME committee as they don't seem to have gotten the memo:

(a) The surface or surfaces of a regular feature of size shall not extend beyond a boundary (envelope) of perfect form at MMC.

The statement was made in an attempt to rein in the discussion for alignment with the terms defined in Y14.5. Now that was "useless".

The discussion has moved forward with the understanding that we are attempting to connect the concepts of true profile and AME in a way which is not strictly contained in the standard. OP (Burunduk) has indicated a desire to continue this direction of discussion.

Your objection is noted.

 

[chez311]

A conical simulator for a secondary datum feature will probably need to expand or contract or progress from MMB to LMB (for external feature) in a way equivalent to how it is shown in fig. 4-31 illustration (a) for a planar feature, which in my opinion could be added to the examples of "the second type of behavior" by pylfrm.

I am largely in agreement with this. This would be indicative of the behavior of a RAME and is the reason a nonFOS as shown in 4-31(a) can have an MMB.

As a result could one make an argument that when referenced with appropriate datum references (ie: sufficiently constrained), a cone or planar feature is an IFOS? Perhaps, seeing as the actual definition in Y14.5-2009 only specifies that a IFOS can contain/be contained by an AME - it does not say which type. In Y14.5-2018 this has been changed to explicitly state that an IFOS can contain/be contained by a UAME.

1.3.32.2 Irregular Feature of Size
irregular feature of size: the two types of irregular features of size are as follows:
(a) a directly toleranced feature or collection of features that may contain or be contained by an actual mating envelope that is a sphere, cylinder, or pair of parallel planes
(b) a directly toleranced feature or collection of features that may contain or be contained by an actual mating envelope other than a sphere, cylinder, or pair of parallel planes

3.35.1 Irregular Feature of Size
irregular feature of size: there are two types of irregular features of size, as follows:
(a) a directly toleranced feature or collection of features that may contain or be contained by an unrelated AME that is a sphere, cylinder, or pair of parallel planes. See Figure 7-41.
(b) a directly toleranced feature or collection of features that may contain or be contained by an unrelated AME other than a sphere, cylinder, or pair of parallel planes. See Figures 7-40 and 11-29.

Regardless a conical/planar primary datum feature or feature without sufficient constraint (due to datum references and precedence) does not exhibit this second type of behavior, at least not without unlimited offset.

But, this is what the standard we work by prescribes, and - for a conical simulator I would say "maximum possible contact" or a term I personally would prefer - stable contact is less ambiguous than in the planar case, because of the centering and effect of cone-to-cone interaction.

I think you'll find that the amount of stability and rocking between a planar and conical feature with convex error is extremely similar. The latter may seem more stable because of the effects of friction, gravity, and compliance of both part and fixture/simulator - if you have a part which is light enough (and with a low enough friction coefficient) to not wedge itself into the simulator or whose simulator is oriented in such a way that gravity does not aid the wedging action (does not act along the simulator axis in the direction of the material ie: a simulator oriented sideways with the simulator axis parallel to the ground) or even a part whose mass is not symmetrically distributed around the conical feature of interest this instability will be more apparent.

When the geometry is considered the behavior (rocking/instability) of a conical feature is very similar to that of a planar feature with convex error.

 

[chez311]

In general, determination of a feature's UAME (if it exists) or RAME(s) doesn't have anything to do with the tolerances that are (or are not) applied to that feature.

This was the key point that I was missing, I got tripped up a bit thinking they might be one and the same or possibly coincident (ie: true profile which defines the tolerance zone and true profile used as the basis of uniform offset for a UAME/RAME).

What do you think about a definition per Y14.5-2009 (per my 5 Dec 19 01:31 post 2009 only specifies AME, changed to UAME in 2018) of a feature as an IFOSb with appropriate datum references which is sufficiently constrained so that a RAME can be established?

 

[pylfrm]

If the position tolerance in Fig. 4-45 referenced [box]B[/box][box]A[/box] instead of [box]A [x,y,u,v][/box][box]B [z][/box], then the conical surface B would be an example if this as well.

This should have said "conical surface A". Sorry about that.

A conical simulator for a secondary datum feature will probably need to expand or contract or progress from MMB to LMB

Not always. Imagine a modified version of ASME Y14.5-2009 Fig. 4-44 where the large cylindrical surface is identified as datum feature B, and the position tolerance references [box]B[/box][box]A[/box] instead of [box]A[/box]. In this case the datum feature simulator would be fixed coincident with the true profile. Progressing toward the material is not viable because there is nothing to limit the progression. Also in this case progression would not change anything about the geometry of the datum feature simulators or the relationship between them.

fig. 4-31 illustration (a) for a planar feature, which in my opinion could be added to the examples of "the second type of behavior" by pylfrm. If for the part in that figure we would rely on the ability of the actual feature to limit the movement of the simulator, then there too unlimited offset of the simulator from its initial position would occur, as the part would rotate around the primary axis.

I think ASME Y14.5-2009 Fig. 4-31(a) belongs in a third category because the datum feature is unable to provide a limit for the offset of the simulator.

1. The range of translation of the simulator is pre-limited from the MMB to the LMB, there would be no purpose in crossing these limits - don't see why this can't also be the case for a secondary conical datum feature (this echoes my previous attempts to convey the need in replacing "true profile" with MMB as the origin of the offset in question).

What happens when the datum feature does not have a profile tolerance with the specific datum feature reference(s) necessary to define those limits? What if it doesn't have any tolerance at all, as is common on parts machined from castings defined by separate drawings?

What happens when the datum feature has the tolerance necessary to define those limits, but does not meet that tolerance?

Regarding MMB my assertion is that for any closed surface (including conical) defined with basic dimensions and a profile tolerance a maximum material boundary exists (I.e. outer boundary for an external feature). This alone already makes "unlimited offset" impossible.

A cone is not a closed surface, at least according to the only definition I'm familiar with. I'm not sure what you mean there.

2. The simulation is required to stop when "maximum possible contact" is made. I don't like the concept of maximum possible contact either - if the surface of datum feature B is produced as a convex arc (within its planar profile tolerance range) then maximum possible contact is pretty much ambiguous.

It doesn't take any special as-produced geometry to show that "maximum possible contact" is not a useful definition. Essentially any ordinary imperfections will do the job.

Fortunately, ASME Y14.5-2009 para. 4.11.4 subparas. (a), (b), and (c) imply that "maximum possible contact" might be intended to mean maximum possible progression toward the feature limited by contact with the feature. This interpretation avoids most of the ambiguity and matches the actual mating envelope definitions.

for a conical simulator I would say "maximum possible contact" or a term I personally would prefer - stable contact is less ambiguous than in the planar case, because of the centering and effect of cone-to-cone interaction. This further rejects any type of "unlimited offset" during simulation.

In cases like the modified version of Fig. 4-44 I described above, different offsets do not lead to different contact.

Then, if it is still the primary datum feature simulation / UAME simulation under question, it seems that "unlimited offset" issue is eliminated in that regard:

Per the above, for a conical feature, there is no issue with the simulation of the type of AME that defines an irregular feature of size.

As I've tried to explain, I don't see any connection between actual mating envelopes and conical (or planar) primary datum feature references. It's two very different types of behavior.

What do you think about a definition per Y14.5-2009 (per my 5 Dec 19 01:31 post 2009 only specifies AME, changed to UAME in 2018) of a feature as an IFOSb with appropriate datum references which is sufficiently constrained so that a RAME can be established?

I addressed this to some extent in the second paragraph of my 4 Dec 19 04:20 post. I think the failure to specify the type of AME was basically an editorial error.

In the end I'm not sure it matters much. I can't think of a case where it would be necessary to determine whether or not something is an irregular feature of size in order to properly interpret a drawing.


pylfrm

 

[Burunduk]

chez311,
That's an interesting note about the change in the definition of IFOS, now required to contain or be contained by specifically an unrelated AME. Then we should keep analyzing conical UAME, as it seems that there are still contradicting statements being made:

Regardless a conical/planar primary datum feature or feature without sufficient constraint (due to datum references and precedence) does not exhibit this second type of behavior, at least not without unlimited offset.

The point is that planar primary datum feature references don't behave in that nonsensical way, and neither do conical primary datum feature references.

You can see in pylfrm's original message that "don't behave in that nonsensical way" is in the context of unlimited offset. Naturally, I side with pylfrm's statement. My explanation to this, however, is probably different than pylfrm's (as he doesn't seem to share my opinion of the un-involvement of true profile in the UAME concept). Here is another attempt to support my position: a primary datum feature simulator is not constrained to a DRF and doesn't "acknowledge" any theoretical boundaries/envelopes that are fixed in location; it "recognizes" in space nothing but the feature that it simulates. Therefore measurement of offset, limited or unlimited, to a fixed (relative to what?) "true profile" boundary/envelope is not required, and if an offset to such third-party envelope is measured anyway, there is nothing that imposes that this offset is expected to accumulate during simulation (because as noted, that reference envelope has nothing to be fixed to).

About the issue of potential rocking/instability of a cone at the interaction with a simulator - similar behavior will be observed when an undisputed feature of size such as a cylinder or width-type feature is produced with a convex error. Most features, regardless of type, are expected to experience the analogical consequences of form error. In the context of the analogy between a cone and a cylinder - it can be noted that a 1 degree included angle cone will be almost identical in all aspects to a cylinder. I generally don't find it very constructive to discuss this type of extremes but since you brought up the 179 degrees cone earlier in the thread - there is your counterexample.

 

[Burunduk]

A conical simulator for a secondary datum feature will probably need to expand or contract or progress from MMB to LMB

Not always. Imagine a modified version of ASME Y14.5-2009 Fig. 4-44 where the large cylindrical surface is identified as datum feature B, and the position tolerance references |B|A| instead of A.

Since the referenced discussion is of secondary conical datum features, did you mean |A|B| reference for position? And, what is datum feature A in the modified case? Is it the cylindrical shank or the flat shoulder?
Sorry for that, from some reason I read conical instead of cylindrical. I agree that in this case the simulator should not be required to progress, as there is no basic location constraint other than coaxiality.

As I've tried to explain, I don't see any connection between actual mating envelopes and conical (or planar) primary datum feature references. It's two very different types of behavior.

I don't understand. If you don't find the primary datum feature simulator and UAME simulator behavior equivalent, why did you bring up all those figures from chapter 4 (Datum Reference Frames) as examples after being asked to address your proposed concept of conical UAME being required to "progress normal to a true profile"? EDIT FOR CLARIFICATION: This is certainly not intended as criticism as I am thankful for the effort and it was beneficial for me to go over all these examples again, but I am still trying to understand the logic of this: Unrelated actual mating envelope simulation for a conical feature can't work, because uniform offset from true profile during simulation becomes unlimited. This is while uniform offset is required for datum feature simulation, which isn't related at all to actual mating envelopes. In addition, primary datum features, tapered or planar, are not subject to this unlimited offset. The more I try, the farther I am from succeeding at making any conclusion out of this.

Additional edit because I wasn't able to address all points at the time of submitting the post:

What happens when the datum feature does not have a profile tolerance with the specific datum feature reference(s) necessary to define those limits?

What if it doesn't have any tolerance at all, as is common on parts machined from castings defined by separate drawings?

What happens when the datum feature has the tolerance necessary to define those limits but does not meet that tolerance?

If a conical datum feature does not have a profile tolerance with the datum feature reference needed to define MMB/LMB, this may withhold the required "progressing" behavior of a datum feature simulator which is prescribed a specific location relative to a higher precedence datum and referenced RMB, regardless of the cone being a FOS or not.

If the feature doesn't have any tolerance at all on the drawing and used as a datum feature, the drawing should direct to another drawing where the tolerance is specified.

If the feature doesn't meet the tolerance then the part is either rejected or corrected before simulating the datum feature.

A cone is not a closed surface, at least according to the only definition I'm familiar with. I'm not sure what you mean there.

That is true, I should have described it as a surface of revolution instead.

I think ASME Y14.5-2009 Fig. 4-31(a) belongs in a third category because the datum feature is unable to provide a limit for the offset of the simulator.

And yet this feature is a valid datum feature and not any less of a feature of its type (planar). Then perhaps too much emphasis is put in this and related threads on the ability of a feature to limit movement of simulators and offsets and it really doesn't have that much to do with the classification of features to types? That is notwithstanding my assertion that the surface of a conical feature DOES limit an actual mating envelope simulator just like it limits a datum feature simulator

 

[dtmbiz]

Chez311
I did intend my statement regarding you as a compliment. I used (usually) because as in one of your posts which I have not dimissed;

I see I may have fired that one off before thinking about it enough. We were talking about establishing a UAME therefore it would indeed be unconstrained in location/orientation. I think some wires may have gotten crossed there.

The “thing” about written communications is that “the tone” isn’t always received as intended.

I do believe there are distinct differences in the terms mentioned.

Boundary is a calculable zone determined by specifications for feature definition
Envelope requires the measurement or contact of imperfectly produced features (in context of Y14.5)
In Rule #1 (2.7.1 para. (a) Boundary and Envelope are coincident.

Despite your continual insistence that there is a concrete and exacting definition for such terms only one of those three is explicitly defined in the standard - this is "true profile". I've already repented for and retracted my statement several times referring to the true profile as a boundary.

My posted statements are not “continual insistence” they are consistent reiteration of my interpretations of Y14.5. They are not “about you” specifically.

The second two are only defined contextually with other concepts ie: Maximum Material Boundary or Unrelated Actual Mating Envelope. If you believe the words boundary and envelope have individual, special, or sacred meaning you should probably alert the ASME committee as they don't seem to have gotten the memo:

I do believe they have distinct meanings and so does Merriam – Webster. Those meanings can “overlap” at times.
Boundary:

something that indicates or fixes a limit or extent

Envelope:

2) something that envelops : WRAPPER

Your objection is noted.

I did not object, it was an attempt to “focus” as my preference.

I find nothing "sacred" or worthy of repentence about Y14.5 or this forum.
Since you would like to give me "action items", e.g."...alert the ASME committee..." let me give you some friendly and unsolicited advice";
'Check' your emotions before you cast insults.

 

[chez311]

You can see in pylfrm's original message that "don't behave in that nonsensical way" is in the context of unlimited offset. Naturally, I side with pylfrm's statement.

I agree with pylfrm's statement you referenced, however I think some context is necessary. Planar/conical primary datum features don't behave in that way because they are simulated by the first type of behavior (envelope coincident to the true profile). The second type of behavior (envelope offset from the true profile) which would be necessary for the definition of a UAME is not a viable option for planar/conical primary datum features because of this "nonsensical" unlimited offset in those cases. This first type of behavior does not result in a UAME, therefore the simulation of a conical/planar primary datum feature does not involve a UAME:

I'd say this second type of behavior involves the defining characteristics of an AME, or UAME in the special case of a primary datum feature reference. The first type of behavior does not involve these characteristics, but it's the only viable option for a conical primary datum feature reference.

a primary datum feature simulator is not constrained to a DRF and doesn't "acknowledge" any theoretical boundaries/envelopes that are fixed in location; it "recognizes" in space nothing but the feature that it simulates. Therefore measurement of offset, limited or unlimited, to a fixed (relative to what?) "true profile" boundary/envelope is not required

The true profile of a UAME would not be fixed or constrained by definition, I don't think I said otherwise? I'm not sure I follow why you don't think offset is applicable in the case of a primary datum feature in which the UAME would be the envelope of interest. Take a cylindrical feature defined with basic dimensions called out at RMB as a primary datum feature - the UAME would uniformly offset from the true profile in the direction of the material until it reached a minimum (for an external feature) or a maximum (for an internal feature). This would be the second type of behavior pylfrm mentioned, the fact that this results in "nonsensical" behavior for conical/planar primary datum features (and therefore the first type of behavior is the only viable option in those cases) doesn't mean it doesn't apply to other types of features during primary datum feature simulation.

About the issue of potential rocking/instability of a cone at the interaction with a simulator - similar behavior will be observed when an undisputed feature of size such as a cylinder or width-type feature is produced with a convex error. Most features, regardless of type, are expected to experience the analogical consequences of form error.

Under most circumstances this is not the case as a local minimum or maximum can be found. This was discussed at length in (

https://www.eng-tips.com/viewthread.cfm?qid=450175)

and I showed in my post on (18 Mar 19 16:54) that only with significant form error and a high width to length ratio will such behavior be possible. Convex error in most cases will not result in such behavior. Pylfrm also shows this in the post (19 Mar 19 05:28).

The same cannot be said for planar and convex features - any amount of convex error will result in the potential for rocking and instability.

In the context of the analogy between a cone and a cylinder - it can be noted that a 1 degree included angle cone will be almost identical in all aspects to a cylinder.

I disagree. The behavior of a cone with 1 or 0.1 or 0.001 degree included angle will be very similar to any other cone and fundamentally different than a cylinder, although such differences may be difficult to discern with the naked eye when looking at a physical part.

Lets break down their behavior in context with the concept of uniform offset from the true profile - if we take the radial (normal to the axis) and axial (parallel to the axis) components of the uniform offset to be r_offset and a_offset respectively, a planar feature or cone having 180deg included angle has an r_offset/a_offset = 0, a cone having an included angle between 0 and 180 has an r_offset/a_offset which is some real nonzero number, and a cylindrical feature has an r_offset/a_offset = undefined (ie: if r_offset is some real number and a_offset is zero therefore r_offset/a_offset = undefined). In contrast to the cylindrical case, r_offset/a_offset for a cone having a nonzero included angle (whether that be 1, 0.1, or 0.00001) approaches infinity as a_offset approaches (but never reaches) zero. A cone with a nonzero included angle will always have a nonzero axial component of the uniform offset (a_offset) and a cone having a 0deg included angle is simply a line.

To put it more simply a cone and a cylinder are fundamentally different as the former has an intersection point (vertex) and the latter does not. As the angle of a cone and therefore this axial component (a_offset) becomes very small it may appear to approximate a cylinder and in physical simulation may behave similarly due to the effects of friction and compliance (and indeed no physical part/simulator/fixture/jig/etc.. is perfect and most produced parts will have "conicity" error at some resolution) but geometrically if we consider the ideal/theoretical shapes the two are still fundamentally different.

 

[chez311]

I think the failure to specify the type of AME was basically an editorial error.

I tend to agree.

In the end I'm not sure it matters much. I can't think of a case where it would be necessary to determine whether or not something is an irregular feature of size in order to properly interpret a drawing.

Not that I would advocate it for features of the type we are discussing (conical/planar), but perhaps to utilize a control which only applies to FOS which didn't require derivation of a UAME - such as applying MMC position similar to Y14.5-2009 fig 8-24 ?

I think ASME Y14.5-2009 Fig. 4-31(a) belongs in a third category because the datum feature is unable to provide a limit for the offset of the simulator.

Interesting, I didn't consider this as I think I was stuck yet again thinking of the tolerance zone as a limit to the offset. Wouldn't this third category rely on limiting offset by the ever troublesome "maximum possible contact" ? Maybe contact could be defined by either the candidate datum concept (or minimum separation as specified in 2018) ?

 

[Burunduk]

chez311, I think I understand now better the assertion that UAME simulation needs to be performed according to the "second type of behavior". Thank you for the clarification and for the relevant quotes from pylfrm's post from 3 Dec 19 05:39. Unfortunately, it seems I didn't pay enough attention to this part when originally reading that post.

chez311, pylfrm, is "unlimited offset from true profile" just a way to say that you think that a UAME simulator is unlimited in contraction/expansion by the tapered feature during interaction with it at an attempt to contain it? In other words, is it an attempt to communicate that a car's brakes don't work during a drive by saying "the vehicle is unable to cease from increasing the distance of its location from the origin of its trip"? If yes, why not just say "unlimited contraction/expansion" without involving offsets and a "true profile" which is not part of the UAME definition in Y14.5? What value does true profile add to the discussion of UAME? Isn't it the same value (lack of it, actually) of notifying the origin of the above-mentioned car trip when describing the fact that the brakes stopped working?
If unlimited contraction/expansion is the issue, this can be discussed. But I am still distracted by not understanding the merit of "unlimited offset" and the involvement of true profile. Is my interpretation above correct? I would like to address other points such as the comparison of cones and cylinders vs. cones and planar features, and the UAME envelope stability issues related to form error of different types of features that we started to discuss, but I really would like to understand the essence of the "unlimited offset" concept first. Thank you for your cooperation.

 

[chez311]

is "unlimited offset from true profile" just a way to say that you think that a UAME simulator is unlimited in contraction/expansion by the tapered feature during interaction with it at an attempt to contain it? [...] If yes, why not just say "unlimited contraction/expansion" without involving offsets and a "true profile" which is not part of the UAME definition in Y14.5? What value does true profile add to the discussion of UAME?

I think pylfrm addressed this previously in this thread:

I think it's best to avoid the expansion/contraction terminology. To me it implies a change in geometry, but a surface with a uniform offset* from a cone is just an identical cone. Same with a plane.

*Yes, I did (and do) mean normal to the true profile.

The notion of expansion/contraction also breaks down for features which are aren't purely external or purely internal, such as the interlocking profiles on the ends of this floor mat segment:

Expansion/contraction gives us a good idea of how the envelope is supposed to behave intuitively, however the terms "uniform offset" and "true profile" can be unambiguously applied to any geometry and is in my opinion more precise. I don't think theres anything inherently wrong with utilizing expansion/contraction if it makes more sense to you in those terms, I just think its less clear in some cases.

Edit: In all fairness, I have utilized the terms in conjunction and interchangeably even previously in this thread, I just gravitated away from it in an attempt to be more consistent.

In regards to a conical/tapered boundary expanded/contracted against (or uniformly offset toward) a conical/tapered feature unconstrained to any DRF it will allow infinite expansion/contraction (uniform offset). A conical/tapered boundary will not contain a conical/tapered feature.

 

[Burunduk]

Expansion/contraction gives us a good idea of how the envelope is supposed to behave intuitively, however the terms "uniform offset" and "true profile" can be unambiguously applied to any geometry and is in my opinion more precise.

Those terms cannot be applied unambiguously to the concept of UAME, as they are not part of the standard's definition of that concept, and the "true profile" which you use in that context is not the same true profile which is defined by the standard. As you have seen, the standardized true profile only defines tolerance zones for a profile control and is undefined without basic dimensions. You had to do guesswork and ask questions to get clarifications from pylfrm with altered terms ("non-literal interpretation" of the standard), this is not what I would call "unambiguous" and "precise".

I think it's best to avoid the expansion/contraction terminology. To me it implies a change in geometry, but a surface with a uniform offset* from a cone is just an identical cone. Same with a plane.

This is a statement I don't understand from two reasons:
1. How does expansion/contraction imply a change in geometry (other than size)? A cylindrical envelope doesn't become an envelope of square cross-section after contraction. "A surface with a uniform offset from a cone is just another cone...". Sure, what else can it be? If we simulate an external cone with a certain base diameter and truncation diameter and a certain included angle, the simulator can begin with larger diameters and contract, while keeping the basically defined included angle, until contact that surrounds the feature. Yes, it stays a cone all the way.
2. Expansion/contraction terminology should be avoided? Definitely not if you communicate with people who follow the standard. It's the only terminology that describes "the second type of behavior" AND is part of the UAME definition in the Y14.5 standard. See the reminder below.

Reminder:

"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)."

Initially, I thought that an adjustable simulator that satisfies the described "second type of behavior" shouldn't be used for a cone, but it was you who insisted otherwise:

Skipping over whether a UAME exists, yes I've already said several times that a fixed conical simulator could be used in place of an "expanding" one - thats not to say an "expanding" one (or one with uniform offset) is not possible.

Can a tapered feature limit the expansion/contraction of a UAME envelope? Yes, it can. If we look at an external conical feature, an internal conical simulator can contract around it radially (which effectively will be the same as "normal to the true profile"), until it contacts it in a way that any additional contraction will result in axial relative movement, irrelevant to the simulation process. At this state there will be enough contact between the simulator and the feature to satisfy both the requirements of Y14.5-2009 (quoted above) AND the mathematical definition:

1.4.13 Envelope, actual mating. A surface, or pair of parallel surfaces, of perfect form, which correspond to an actual part feature of size, as follows:
(a) For an External Feature. A similar perfect feature counterpart of smallest size which can be circumscribed about the feature so that it just contacts the surface.
(b) For an Internal Feature. A similar perfect feature counterpart of largest size which can be inscribed within the feature so that it just contacts the surface.

As for cones vs. planar features vs. FOS

As pylfrm postulated in the originally referenced thread, do you not see some strong similarities of a conical feature having 179deg included angle and a planar feature? In fact I think an argument could be made that the distinction between a planar feature and a conical/tapered feature having 180deg included angle is more philosophical than mathematic - geometrically the two are indistinguishable.

The behavior of a cone with 1 or 0.1 or 0.001 degree included angle will be very similar to any other cone and fundamentally different than a cylinder

A 1 degree included angle cone does not differ from a cylinder more than a 179 included angle cone differs from a planar surface. A 179 cone is still a surface of revolution with a defined axis of non-zero length. Actually, in that sense, even a 179 degrees cone has more in common with a cylinder than with a planar surface, hence not as detached from the FOS category as you portray it to be.

I still can't say with confidence that tapered features are features of size, but I am not convinced by the arguments presented so far why they aren't.

 

[3DDave]

Axial displacement is all that is required for one cone to fit another. There is no "expansion" required. On that basis there is no fundamental size to a cone. This is similar to normal displacement to fit one plane to another.

The point of FOS prior to '2009 was to give a single value to such features to allow adding and subtracting tolerance and variation values for conformance purposes. By making a definition that allows all features to be FOS there is no longer any particular meaning to the term at all.

At this I'll leave off discussing the defective definitions that were introduced in the '2009 version to justify not having to correctly describe the analysis of non-FOS volumetric features.

 

[Burunduk]

Axial displacement is all that is required for one cone to fit another

Perhaps, but an adjustable simulator can be used too.

At this I'll leave off discussing the defective definitions that were introduced in the '2009 version to justify not having to correctly describe the analysis of non-FOS volumetric features.

It was probably desired to allow applying MMC/LMC controls (surface interpretation/boundary control) to features other than cylinders and tabs. Good thing? Bad thing? Don't know, it is what we have. Probably someone finds it useful.

 

[pylfrm]

chez311, pylfrm, is "unlimited offset from true profile" just a way to say that you think that a UAME simulator is unlimited in contraction/expansion by the tapered feature during interaction with it at an attempt to contain it?

I don't think "contraction/expansion" or "smallest/largest" are meaningful concepts for a cone considered in isolation. I'd say the same about planes and various other shapes, such as the one mentioned in my 15 Nov 19 04:05 post.

ASME Y14.5-2009 para. 1.3.25.1 says a UAME is "a similar perfect feature(s) counterpart". I take this to mean that the geometry of the UAME has some definite relationship to the "perfect feature" geometry. The only relationship I've thought of that makes any sense is uniform normal offset, and this seems to match the standard's descriptions of UAME behavior. As for the "perfect feature" part, this is something that must be defined by the drawing. If the drawing fully defines the true profile of the feature, I think it would be hard to argue that the perfect feature geometry is anything else.

Continuing with this reasoning and considering the specific example of an external cylindrical feature, "envelope with a uniform offset from the true profile" describes an infinite family of cylindrical envelopes. The only purpose of the true profile here is to fully define this family, and that can be achieved without the true profile itself being fully defined. In this case, "cylinder of unspecified diameter" is sufficient. One particular member of this family is the unrelated actual mating envelope, and that is the smallest one that can contain the feature while remaining outside its material.

For a conical envelope, "smallest" and "largest" are not really meaningful, but I think the concept can be reasonably generalized to "maximizing the offset in the direction toward the material while remaining outside the material" (or something like that; I have a hard time phrasing this concisely). This works fine for a RAME with appropriate higher-precedence datum feature references because different amounts of offset produce conical envelopes with different apex locations, even though they are all congruent. The fact that the envelopes are all congruent is problematic for a UAME because there is no way to distinguish between them.

I'd be interested to hear other opinions about what "a similar perfect feature(s) counterpart" means.

1. How does expansion/contraction imply a change in geometry (other than size)?

A uniform normal offset from a cylindrical surface of a certain diameter produces a cylindrical surface of a different diameter. A uniform normal offset from a conical surface produces an identical conical surface, just axially displaced.

I'd say a change in size is a change in geometry. An axial displacement is not, at least when we're talking about the surface itself and not its relationship to anything else.

If we simulate an external cone with a certain base diameter and truncation diameter and a certain included angle, the simulator can begin with larger diameters and contract, while keeping the basically defined included angle, until contact that surrounds the feature.

It sounds like you're talking about changing the diameters of circles that define the extent of a conical surface, not changing anything about the conical surface itself.

The extent of a datum feature simulator should be sufficient to cover the entire datum feature. Beyond that, extent is irrelevant. That's why it's simpler to consider simulators as having infinite extent.

2. Expansion/contraction terminology should be avoided? Definitely not if you communicate with people who follow the standard. It's the only terminology that describes "the second type of behavior" AND is part of the UAME definition in the Y14.5 standard.

The meaning of expansion or contraction when applied to a cylindrical or spherical surface is fairly obvious--the diameter increases or decreases. It doesn't really matter whether this is achieved by uniform normal offset or isotropic scaling or some other transformation.

The meaning of expansion or contraction when applied to a conical surface is unclear. A uniform normal offset or isotropic scaling doesn't change anything about the geometry of the surface, so I wouldn't consider that expansion or contraction. The only thing I can think of that might deserve the name is an increase or decrease of the included angle, but that doesn't make sense when we're talking about features defined with basic angles.

Can a tapered feature limit the expansion/contraction of a UAME envelope? Yes, it can. If we look at an external conical feature, an internal conical simulator can contract around it radially (which effectively will be the same as "normal to the true profile"), until it contacts it in a way that any additional contraction will result in axial relative movement, irrelevant to the simulation process.

If the "contraction" you describe does not change the geometry of the envelope, how can any particular amount of contraction produce a different result than any other amount?


pylfrm

 

[chez311]

How does expansion/contraction imply a change in geometry (other than size)?

I believe the change in size is what is referred to. A cylindrical feature (controlled at RFS/RMB) which has a true profile of a 10mm dia cylinder but is produced at 15mm dia cannot be simulated by a cylinder of the same exact geometry as the true profile of 10mm dia - it MUST expand and increase in size to 15mm dia. This is in contrast to a conical feature which no matter how it is produced can be simulated by a conical simulator of the same exact geometry (same included angle) as the true profile, no change is necessary.

Expansion/contraction terminology should be avoided? Definitely not if you communicate with people who follow the standard.

This is a fair point. I'm not going to speak for pylfrm as there may be differing opinions on this, but unless the verbiage in the standard were changed to reflect this if I were conversing with others about undisputed RFOS or similar standard cases I don't think I'd introduce the concepts of "uniform offset to the true profile" unless there was some specific need - thats not to say its not applicable in those cases, I just don't think its necessary to introduce in clear cut cases.

Initially, I thought that an adjustable simulator that satisfies the described "second type of behavior" shouldn't be used for a cone, but it was you who insisted otherwise:

Thats not what I was saying. Is it possible to create such a simulator? Yes. Does such a simulator behave in such a way to allow unlimited expansion/uniform offset? Also yes. Perhaps I shouldn't have added that clarification, my insistence was to try and conjure the image of a physical expanding(offsetting) simulator which one could expand without limit as an example of what was meant by unlimited offset - this seems to have created some confusion.

Can a tapered feature limit the expansion/contraction of a UAME envelope? Yes, it can. If we look at an external conical feature, an internal conical simulator can contract around it radially (which effectively will be the same as "normal to the true profile"), until it contacts it in a way that any additional contraction will result in axial relative movement, irrelevant to the simulation process.

This is essentially the first type of behavior and is only valid if the envelope is coincident with the true profile. Even if whatever one determines as the ideal "amount" of contact is achieved this does not provide a limit to the offset for a conical feature - if the envelope is allowed to offset relative to the true profile it may continue to offset infinitely from the true profile without changing the "amount" of contact, likewise the "amount" of contact can be infinitely changed/adjusted without changing the amount of offset. For a conical feature and an envelope unconstrained to any DRF (ie: UAME) the amount of offset is wholly separate from the "amount" of contact and therefore we could say any amount of expansion/offset is irrelevant to the simulation process, which is why the first type of behavior is the only one available to a conical primary datum feature.

A 1 degree included angle cone does not differ from a cylinder more than a 179 included angle cone differs from a planar surface. A 179 cone is still a surface of revolution with a defined axis of non-zero length. Actually, in that sense, even a 179 degrees cone has more in common with a cylinder than with a planar surface, hence not as detached from the FOS category as you portray it to be.

I would say it does. A cone with a 1 degree included angle as well as one with 179 degrees and 180 degrees have an intersection point (vertex). This is also what results in the nonzero axial component for a cone of any nonzero angle - including 180 degrees. The fact that it is not obvious where this intersection point is for a 180 degree included angle cone does not mean one cannot be found - indeed an infinite number can be found. A cylinder has exactly ZERO intersection points (vertices) and therefore has ZERO axial component - neither of these is possible for a cone of any nonzero angle (1, 179, and 180 degrees included). This is the fundamental difference to which I referred.

 

[Burunduk]

For a conical envelope, "smallest" and "largest" are not really meaningful

...

The extent of a datum feature simulator should be sufficient to cover the entire datum feature. Beyond that, extent is irrelevant. That's why it's simpler to consider simulators as having infinite extent.

A cylindrical feature (controlled at RFS/RMB) which has a true profile of a 10mm dia cylinder but is produced at 15mm dia cannot be simulated by a cylinder of the same exact geometry as the true profile of 10mm dia - it MUST expand and increase in size to 15mm dia. This is in contrast to a conical feature which no matter how it is produced can be simulated by a conical simulator of the same exact geometry (same included angle) as the true profile, no change is necessary.

...

This is essentially the first type of behavior and is only valid if the envelope is coincident with the true profile.

Y14.5 discusses mainly theoretical envelopes and theoretical datum feature simulators. A theoretical conical envelope can indeed be infinite and successfully contain (yes, I think the word "contain" is appropriate) any produced feature within its tolerance zone, as long as the included angle of the simulator is the basic angle specified. However, I think that the intent of the standard is to allow the use of physical datum feature simulators/envelope simulating devices for all cases. A physical simulator cannot be infinite, and to accommodate it, even for a theoretical envelope it isn't MUST to be infinite. It can also be truncated as long as the truncation diameter is smaller than the truncation diameter of the feature (as produced). Arguably a theoretical envelope with an initial base diameter and truncation diameter can contract/expand until meeting the feature - all local circular elements along the finite envelope will increase or decrease in size proportionally. In physical reality a simulator that behaves this way - essentially expands/contracts as specified by the UAME definition, is sometimes the only practical solution; imagine a conical feature connecting to a flat shoulder of a larger cylinder at its smaller end. Since there is no way to mate a conical simulator axially to this feature, A physical primary datum feature simulator must be a chuck-like device that will contract around the feature, obviously also representing the UAME. Isn't it "the second type of behavior"?

The fact that is not obvious where this intersection point is for a 180 degree included angle cone does not mean one cannot be found - indeed an infinite number can be found.

Since we are talking about theoretical geometry, a 180 degrees cone is perfectly flat, and all line elements on it are parallel and coplanar. Parallel line elements do not intersect, therefore there is no intersection point/vertex.

 

[pylfrm]

A physical simulator cannot be infinite, and to accommodate it, even for a theoretical envelope it isn't MUST to be infinite. It can also be truncated as long as the truncation diameter is smaller than the truncation diameter of the feature (as produced).

I agree, but any truncation is only an implementation detail and is not relevant to the meaning of tolerances or datum feature references or determination of UAMEs or anything like that.

Arguably a theoretical envelope with an initial base diameter and truncation diameter can contract/expand until meeting the feature - all local circular elements along the finite envelope will increase or decrease in size proportionally.

I assume you have in mind a radial contraction or expansion that applies not only to the surface of the envelope but also to the circles bounding its extent. If not, please clarify.

Instead of saying the envelope contracts or expands, wouldn't it be at least equally valid to say that the circles bounding the extent of the envelope move toward or away from the apex?

Regardless of what we call this, I don't know what you mean by "until meeting the feature" if we are talking about a UAME or primary datum feature reference here. The feature can be brought into contact with the envelope before any contraction or expansion takes place.

In physical reality a simulator that behaves this way - essentially expands/contracts as specified by the UAME definition, is sometimes the only practical solution; imagine a conical feature connecting to a flat shoulder of a larger cylinder at its smaller end. Since there is no way to mate a conical simulator axially to this feature, A physical primary datum feature simulator must be a chuck-like device that will contract around the feature, obviously also representing the UAME.

It may be more practical to have a fixed conical socket which can be temporarily split in half to allow installation of the part. Sacrificing contact with some small portion of the surface near the small end would probably be at least as acceptable as sacrificing contact with the portion of the surface between the chuck jaws.

Isn't it "the second type of behavior"?

It is not, assuming we are talking about a primary datum feature reference. The datum feature simulator would remain coincident with the true profile regardless of whether the simulator is realized as a chuck-like device or a splittable socket. To put it in terms of datums instead of true profile, the datum point created would be at the apex of the simulator in its final configuration.


pylfrm

 

[Burunduk]

I assume you have in mind a radial contraction or expansion that applies not only to the surface of the envelope but also to the circles bounding its extent. If not, please clarify.

Yes, it is what I meant.

Instead of saying the envelope contracts or expands, wouldn't it be at least equally valid to say that the circles bounding the extent of the envelope move toward or away from the apex?

I would say that the bounding circles moving away or toward an apex can be relevant for a theoretical envelope, but this description is more detached from physical reality than the change in size of circular elements. Everything should be able to eventually be simulated physically, and physical devices don't work this way.

Regardless of what we call this, I don't know what you mean by "until meeting the feature" if we are talking about a UAME or primary datum feature reference here. The feature can be brought into contact with the envelope before any contraction or expansion takes place.

A better wording would be "until it mates to the feature" - mating in the conical case can be performed by the means of axial mating (if possible) or the split conical socket which you mentioned, but an expanding/contracting device is also an option - which satisfies the UAME definition.

The datum feature simulator would remain coincident with the true profile regardless of whether the simulator is realized as a chuck-like device or a splittable socket.

As usual, I have a hard time with this type of assertions. If a true profile of a cylindrical +/- toleranced feature is a cylinder of unspecified size, for sure the true profile can stay coincident with the cylindrical simulator during the entire process as well. Does it matter?

To put it in terms of datums instead of true profile, the datum point created would be at the apex of the simulator in its final configuration.

And, the datum axis of a cylindrical datum feature is always at the center of the simulator.