Composite Frame for Single hole 4

[Madhu454]

Hello,

Please see the attached drawing.
Figure-1 : A composite Position frame is applied for the hole.
Figure-2 : The same hole uses Position control and a perpendicularity control.

Question : 1) Can Composite position frame be applied to single hole.?
2) I have a doubt that,composition position frame is used only when there is a pattern of holes, as ASME says the upper frame 'PLTZF' and lower frame is 'FRTZF'.
3) Is figure-1 is legal specification? or do I need to dimesion the part as shown in Figure-2. or both is Fine?

Plz suggest.


Madhusudhan Veerappa
Mechanical Engineer

 

[ptruitt]

What about this?

Peter Truitt
Minnesota

 

[Belanger]

I hope I didn't sound like a bad sport

I just can't wrap my head around the belief that the word "position" doesn't have to involve a distance. If that statement were true, it wouldn't be position, it would be perpendicularity, parallelism, flatness, straightness, angularity, circularity, cylindricity....

It's true that I was making an absolute statement, but it was one not based on my interpretation of Y14.5, but rather on the very meaning of the word "location."


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

 

[axym]

J-P,

I don't hate to sound like I'm splitting hairs - maybe that's my problem ;^). I guess I've found that applying Y14.5 GD&T in quality and inspection is all about splitting hairs.

Here's my take on your Parallelism/Flatness example. I agree with you that specifying Parallelism when only form control is needed is not OK - it would impose an unnecessary orientation control. But I think a better comparison would be to substitute Surface Profile instead of Parallelism. Surface Profile is capable of controlling size, form, orientation, and location. If applied to a single planar surface, with no datum feature references, Surface Profile would control form only. So would it be legal to do that? I would say yes, but I would still recommend specifying Flatness.

pmarc,

Whether it violates the rule depends on exactly what a pattern is. Here's an excerpt from Section 5.1.2 of the Y14.5.1M-1994 mathematical definitions standard:

"For the purposes of this Standard all tolerances of location are considered to apply to patterns of features, where a pattern may consist of only a single feature."

Again, I would say that applying composite Position to a single feature does not add any value and would introduce confusion. Having said that, composite Position is often applied in other situations where it does not add any value, even in the Y14.5 standard itself. Take Fig. 5-19 in Y14.5M-1994. The FRTZF's only reference datum A, which is orthogonal to the considered features and therefore can only control orientation and not location. So the special rule for composite FRTZF's, in which any datum feature references control orientation only, has no effect. The composite FCF's are functionally equivalent to multiple single-segment FCF's - so why use composite FCF's? The only other difference would be the other special rule with composite FRTZF's, that the rule of simultaneous requirements does not apply. This is not mentioned in the discussion of the figure.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

(Before someone says it: I know that the form symbols listed in my previous post are not controlled by position. I was just throwing out random geometric characteristics!)

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

 

[Belanger]

Evan, I don't know if your example of profile fits the situation being discussed. I agree that profile of a surface can be used to control flatness only. But that's OK because profile tolerances, at their base level, are just form tolerances (that can then elevated to take on variations such as irregular shapes and/or datum references, if desired). But the definition of the word profile doesn't inherently invoke orientation (parallelism), so there's no conflict: leave it stripped to its basest meaning and it is equivalent to flatness.

What I keep trying to say is that position is, at its basest level, a location control, so there is no option of truncating it to do orientation only.

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

 

[axym]

J-P,

Thanks - now you've got me thinking about the semantics and metaphysics of position and location. I've got other things to do, like earning a living ;^).

I hesitate to base anything on the words or categories that Y14.5 assigns to things. I therefore hesitate to judge what Position and Profile inherently invoke, based on the generic meanings of the words "position" and "profile". I prefer to examine the definitions of the tolerance zones and their behavior, and make judgments from there. Exactly what ends up being controlled usually depends on the specific configuration of considered features and datum features.

There may still be a way for both of us to be right:

An orientation tolerance zone is oriented to the datum reference frame, but not located to it. This means that the zone is not allowed to rotate relative to the DRF, but is allowed to translate relative to the DRF. The zone can therefore translate relative to the considered feature.

A Position tolerance zone is oriented and located to the datum reference frame. This means that the zone is not allowed to rotate or translate relative to the DRF. But we can impair its ability to locate the considered feature, by specifying only a single datum feature that is not parallel to the considered feature. This allows the DRF to translate relative to the part, and thus allows the zone to translate relative to the considered feature. This achieves the same result as the orientation tolerance in this special case, but the Position tolerance zone is still located to the DRF.

What do you think?

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Evan ... I was with you until a couple of sentences into the last paragraph. Why are we allowed to "impair" the established meaning of a geometric characteristic symbol? I'm familiar with "upgrading" certain symbols (especially for profile), but it seems to be illogical to downgrade a symbol below its intrinsic definition.

The standard says that the position symbol controls the location of one or more features of size. So to try to downgrade it by removing the location requirement tells me that it's no longer a position symbol.

An automotive analogy: You can buy an old chassis from a junkyard and build onto it, eventually making a driveable car. You can still say that the chassis is there; it's just been embellished.

But to buy a car and strip it down to the chassis only -- throw out the body and the powertrain -- means that you can no longer call it a driveable car.

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

 

[Belanger]

One other thought... The only "impairment" that comes to mind right now is the T modifier, which impairs the surface orientation symbols by removing the form requirement. But that is an explicit modifier made for such an effect; it doesn't damage the intrinsic meaning of parallelism, etc.

If I tried to create a modifier for orientation that removes the angular control back to a datum, then I would be messing with the intrinsic meaning of orientation, and it would be illogical.

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

 

[DeanD3W]

John-Paul,
What I think would be logical is a position tolerance applied to two or more parallel holes, where each references only a planar datum feature that is intended to be perpendicular to the holes... These position tolerances that I speak of here would follow looser position tolerances with more constrained datum reference frames.

The position tolerances I'm speaking of that reference only "A" would have their tolerance zones tied together by simultaneous requirements, so the feature-to-feature relationship and the orientation to "A" would be controlled.

If we pare that back to only one feature, then yes, it is more intuitive to use perpendicularity, but not wrong to use position for that control.

If we must always use the control that provides just the control needed, then as Evan pointed out above, there are fairly common cases where composite position must be replaced with single segment feature control frames.

I think we all agree that perpendicularity is more intuitive for the case being discussed, but the "GD&T police" can issue no citation if position is used instead :^).

Dean

http://www.d3w-engineering.com

 

[dingy2]

J-P:

You put up a valiant battle and have given you a star for it. The other factor is that I totally agree with you on this thought. I somehow have a difficult time placing a positional tolerance on a feature when, in fact, the only relationship is perpendicularity.


Dave D.

http://www.qmsi.ca

 

[ptruitt]

A textbook (GeoTol Pro - 2009) I have says "...standards recognize that that the axis verses surface interpretation can yield minor but different results. ...if there is a conflict...the surface interpretation should prevail."

Peter Truitt
Minnesota

 

[Belanger]

Dean -- you nailed it. The perpendicular-only datum reference is just fine if we're talking about two or more features. But with only one hole, then you gotta have a "distance" datum along with a basic dimension (even if the basic dim is implied).

Thanks, Dave.

My persnicketiness on this is not meant to be an "I'm right, everyone else is wrong" type of thing. I was just driving at what a given word actually means in a dictionary sense.

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

 

[Madhu454]

Hi Jhon-Paul,

I did not get the maning of " The perpendicular-only datum reference is just fine if we're talking about two or more features. But with only one hole, then you gotta have a "distance" datum along with a basic dimension (even if the basic dim is implied)."

Could you please explain this..



Madhusudhan Veerappa
Mechanical Engineer

 

[Belanger]

If two holes are controlled by the same feature control frame, and the position symbol is used, then they are automatically being controlled for location to each other (the distance between them), even if no datum reference is given. It's kind of like each hole becoming a datum for the other.

Dean and I were referring to a similar situation where there is one datum reference (a plane perpendicular to the two holes). This is OK because the datum gives an orientation aspect to the position symbol, but the intrinsic meaning of position -- location -- is still intact because there is a distance between the two holes that is locked.

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

 

[axym]

J-P,

Again, I hesitate to base anything on the labels and categories that Y14.5 uses. Circular Runout is not classified as a tolerance of location - yet it controls location just as much as Concentricity, which is classified as a tolerance of location.

Y14.5 often gives words new meanings that stray far from their common usage. If we try to understand terms used in Y14.5 in a dictionary sense, we're finished ;^). Take "concentricity", "symmetry", "virtual", "resultant", and my all-time favorite "simulate". The list goes on.

It's interesting that you feel that Position is a location control that must do all it is capable of, but Profile is a form control that can get "upgraded". To me, both Position and Profile have the power to locate but this power can be made irrelevant by how the control is applied.

Profile creates a tolerance zone that is completely fixed in its form, size, orientation and location in the datum reference frame. This gives Profile the power to control the form, size, orientation, and location of the considered feature. But we can choose to do one or more of the following things:

-we can make Profile's ability to locate the considered feature irrelevant, by applying it to a single considered feature and specifying only a single datum feature that is orthogonal to the considered feature
-we can make Profile's ability to orient the considered feature irrelevant, by applying it to a single considered feature and specifying no datum features at all
-we can make Profile's ability to control the considered feature's size irrelevant by applying it to a single feature that does not have size i.e. a single planar surface

Position also creates a tolerance zone that is completely fixed in its form, size, orientation, and location in the DRF. This gives Position the power to control the orientation and location of the considered feature. It is not allowed to control form or size, because Position is only applied to resolved geometry (axes, center planes, center points) that by definition have perfect form and zero size. In the same way that we can with Profile, we can choose to make Position's ability to locate the considered feature irrelevant by applying it to a single considered feature and specifying only a single datum feature that is orthogonal to the considered feature.

So I would maintain that applying Position instead of Perpendicularity is definitely not recommended, but no more illegal than applying Profile instead of Perpendicularity or Flatness. For me, the constraints on the tolerance zones speak louder than the names given to them.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Hi Evan,
There are four essential things that GD&T can control: size, form, orientation, and location. Notice that profile is not on that list.

Position is a location control. So guess what -- by definition, it controls location. This has nothing to do with Y14.5; it's black and white simply because it is called a "location" control. It had better control location, or it's not being used correctly. See my chassis example above.

Profile is its own category, but notice that it is not pinned down to size, form, orientation, or location. So this is where we have to look to Y14.5 for how it wishes for profile to be used/interpreted. We see that profile always controls form. So that can never be taken away. But we are allowed to build it up from there.

I see what you are saying about profile, but here's where I disagree: You wrote,
"In the same way that we can with Profile, we can choose to make Position's ability to locate the considered feature irrelevant by applying it to a single considered feature and specifying only a single datum feature that is orthogonal to the considered feature." (my emphasis)

That doesn't make sense. If I claimed that "we can choose to make profile's ability to control form irrelevant," would you see a problem?

Once again, the difference is in:
[li]building up a base definition to add extra qualities to it (which you describe well for profile)[/li]
compared to:
[li]stripping down a base definition to remove constituent qualities (which is what position of a single hole is doing, and therefore negates the meaning of position)[/li]

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

 

[PaulJackson]

It is easy to say "I agree" with out investing all of the time and thought necessary to articulate a response... so said... I agree Evan. Well stated!

I recall defending the position (refuted albeit) that a position of a pattern does not require a datum feature designation... arguing that the control constrains all six dof among the pattern members simply because it is a pattern... only to see evidence of its acceptance and application in the latest version of the standard.

Paul

 

[axym]

J-P,

I'm not sure that Position controlling location by definition is as black and white as you describe. It is definitely something to do with Y14.5, because Y14.5 is where the characteristic was given the name Position and classified as a location control!

Subdividing geometric controls into size, form, orientation, and location is a useful concept, but it's unfortunately not that simple. Despite the categories that Y14.5 divides them into, most of the 14 geometric characteristics control more than one of the four things. Even with form tolerances, which only control one thing, we need to subdivide the idea of form into "local form" of line elements (Straightness and Circularity) and "global form" of the entire feature (Flatness and Cylindricity).

When patterns or groups of features are toleranced together, we also need to subdivide orientation and location into "relative" (to each other) and "absolute" (to the datum reference frame). A Position or Profile tolerance with no datum references controls the relative orientation and location of the features but not their absolute orientation and location. The FRTZF of a composite FCF creates a rigid pattern of tolerance zones that can freely translate relative to the DRF. This could be described as controlling the relative location of the features and not their absolute location.

I don't mind agreeing to disagree on this one. You're able to hang a lot more on the words and names in Y14.5 than I can, and that's fine. The words and names have led me on so many conceptual wild goose chases that I can only look to the geometry to make sense of it all.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Well, true, we do have to look to Y14.5 to see the name "position" and to see that it is a "location control." But I was simply appealing to logic to say that if a location control is stripped of location, then it is not a location control, and thus we have a problem of A = B and A ? B. (If I say parallelism no longer controls orientation, I would hope that you call me on it!)

I have no problem with the "relative" vs. "absolute" designations. Don't know if that applies here, because for both relative and absolute location, the location is always present in some respect.

If I may pick at one more concept given in your post from 13:13, I think this is the crux of the matter:

"Profile creates a tolerance zone that is completely fixed in its form, size, orientation and location in the datum reference frame."

"Position also creates a tolerance zone that is completely fixed in its form, size, orientation, and location in the DRF."

These two statements are not both true. While position's zone is always fixed in form, size, orientation, and location (ignoring bonus tolerance), profile's zone is not. It is always fixed in form only. If you'd like to add parameters to make the other 3 qualities part of profile, then you can do so by adding datums, etc.

That's all I was getting at by my explanation of the "base definition" of something. The stars seem to be lining up on your side so I'm OK to let it go.

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

 

[pmarc]

J-P, a star from me. Not because Evan has two , but first of all because I totally agree with your standpoint (which doesn't mean I don't understand Evan's arguments).