Profile vs Parallel & Perpendicular 1

[SpaciouS]

Is there a difference between the following call-outs?

A double segment FCF that states, for example:
Top segment: Perpendicular tolerance of .02 relative to datum -A-
Bottom segment: Parallel tolerance of .02 relative to datum -B-

and

A single segment FCF, for example:
Profile tolerance of .02 relative to datums -A- & -B-

I see this lot and my feeling is if the double segment FCF doesn't state 2 different tolerances, why not use a single segment profile FCF?


Thanks,
SpaciouS

 

[Belanger]

They are different, because there's no guarantee that datum features A and B will be perpendicular to each other. Assigning both datum references in the single segment implies a hierarchy of datums.

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

 

[drawoh]

SpaciouS,

Parallel and perpendicular tolerances do not control position. All they control is orientation. The profile tolerance is a positional control. Your two callouts do not say the same thing.

JHG

 

[SpaciouS]

Belenger: So if I'm not mistaking, not implying a hierarchy of datums, the tolerance zones are indipendant from each other and are inspected as such? If so, would that mean the potential overall deviation of the specified surface is greater? (I’m working on orientation stacks).

Drawoh: The question I meant to ask, or what I'm trying to figure out is how the tolerance zones differ, regardless of position. Profile does have an effect on "size, form, orientation, and location" (Para 6.5.1 Profile Tolerancing).

Thanks for your responses and I look forward to further discussion.


Thanks,
SpaciouS

 

[drawoh]

SpaciouS,

Let us compare the cases of a profile tolerance of 0.2mm versus a parallel tolerance of 0.,2mm. In either of the cases, a maximum deviation from parallel of 0.2mm is allowed. The perpendicularity specification works the same way.

My point is that profile also controls position.

Belanger's point is that the datums are important. Profile from A|B is not the same as parallel from A and perpendicular from B.

Aside from the fact that A and B may not be nominally perpendicular, there will be some error in the actual condition of your secondary and tertiary datums. The geometry could be weird and complicated.

JHG

 

[SpaciouS]

Datum A & B are nominally perpendicular non-features of size.

Thanks,
SpaciouS

 

[Belanger]

SpaciouS ... Yes, it means that the tolerance zones are independent of one another. See the attached graphic (pic at the top is the print; the other pic is what might really happen). Datums A and B were designed to be made perpendicular to each other, but since there is no direct control on their interrelationship, I've purposely drawn a part that is made with an angle, which might actually happen. The tolerance zone that I show with the two phantom lines fulfills the part about being parallel to A. But to show the tolerance zone for perpendicularity to B, I'd have to draw another pair of phantom lines that is 90º to datum B (which itself is angled!). For the sample part shown, it will probably be impossible to have a part that fits both tolerance zones simultaneously.

If we used a single segment of profile of a surface with A and B, then there would be only one tolerance zone (the one shown in the sketch). Even though the left side is angled, the true datum B would be an imaginary plane that is perfectly perpendicular to A, and touching the left side of the part at only the top edge.

Drawoh, FYI... profile doesn't have to control location (even when datums are referenced). If my parallelism callout were simply replaced with profile of a surface, it would have exactly the same meaning. The location is be controlled by the ± tolerance on the height. If you have a copy of Y14.5, there is an example of this "floating" profile zone in Fig. 8-27 of 2009, or FIg. 170 of the 1994 edition. In those two figures, profile only controls form and orientation.

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

 

[Belanger]

On my last sentence, I meant to refer to Fig. 6-18 of the 1994 standard.

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

 

[ptruitt]

Is it true that, if a drawing were to have all basic dimensions and all profile call-outs with no datums, all profile call-outs would would need to be met simultaneously?

Peter Truitt
Minnesota

 

[drawoh]

Belanger,

I am staring at Fig 8.27 and I am puzzled. Why would you not just specify parallel to datum_A?

I have a tendency to do this the opposite way. I will show a sloppy profile tolerance all around a part, then I will apply an accurate ± tolerance to one feature. Excluding the feature from the profile is a lot of work, and the accurate tolerance is not incompatible with the profile.

JHG

 

[Belanger]

Peter -- I'd like to hear what others think on this one also. But I'd say that all tolerances have to be met simultaneously on every print, unless there is an MMB or LMB modifier invoked. I think that's the only time where "simultaneous" vs. "separate" is an issue.

JHG ... for Fig 8-27, I guess the logic is that parallelism would only speak to that surface's relationship to datum A (datum B couldn't be mentioned if it were a parallelism symbol, since it's not one that's parallel to the surface being toleranced). Because the symbol there is profile of a line -- that's where datum B comes in: datum B gives us a reference plane to orient the imaginary lines or slices across the top that are being toleranced.

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

 

[SpaciouS]

Thanks a lot for all your answers. Belanger I appreciate your extensive explanations, very helpful.

Thanks,
SpaciouS

 

[fsincox]

JP,
Hasn't the standard now accepted the concept of orientation (angularity) to be used in the case of multiple datum references to avoid the "is it parallel vs perpendicular surface issues"?
I agree with the comments above, orientation to a multiple datum framework is not the same orientation specified as separate datum callouts.
Frank

 

[Belanger]

Yep, Frank, I agree too. ANd I think angularity has always been acceptable as the miscellaneous way to control any angle, even parallel and perpendicular at the same time. The new standard is a little more direct about address it however.

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

 

[ptruitt]

Isn't there an opportunity to reduce confusion and reduce the learning curve by using and teaching the various ways that surface profile can be used?

Peter Truitt
Minnesota

 

[fsincox]

I will not disagree that profile could control this, I was taught not to invoke a higher level control if a lower level would do (the KISS priniple).
As some here already know, this is one of my critisims of the standard itself, it necessesarily simplifies things for educational purposes and in doing that leaves us prey to the "it must be like it's shown in the book crowd". It makes the real world look like it fits into neat little boxes. The standard itself states this fact in the begining but people continue to cling to it because of a fear of the unknown.

From an old school manufacturing point of view, "parallel to this" and "perpendicular to that" may make some sense, as it actually may have been produced using (2) separate operations. I was taught to say, "how does it install" and "where do the features have to be in the actual assembly installation". This leads more to more of a framework approach. Like "when mounted on this face and located on the dowel pins the other feature must fall here" or "have some desired orientation".
Profile and compound profile will do this, as noted, but it also may imply locational controls where they are not needed.
I myself, am one of those profile everything kind of guys but it must be done carefully and not just thrown on. It does appear our world is not ready for it yet.
Frank

 

[drawoh]

fsincox,

I do not understand your distinction between lower and higher level specifications. If you specify parallels and perpendiculars, you need more feature control frames. Your drawing is more complicated. Plus/minus tolerances on their own, do not control angles.

If you are determined to control everything, the profile tolerance leads to simpler drawings.

JHG

 

[Belanger]

If you want a surface to be flat, most of us would specify the flatness control. But you can also plug in profile of a surface and it would mean the same thing. But why use the "higher-level" symbol when flatness would be the most direct and understandable way to say it? I think this is what Frank is saying, and it makes sense to me.

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

 

[fsincox]

J-P,
Yes, thank you.

drawoh,
I could not say it is wrong as the standard does not declare it to be so. I just have too many battle scars to recomend it, either. As I stated I am a proponent of profile. I believe a careful reading of the standard's foreword shows they are too, I just do not believe the majority of the world has caught up yet.

ptruitt,
I agree confusion is a serious problem. I have seen proposals to do away with many of the controls based on similar logic that fewer is simpler.

 

[dingy2]

J-P:

You stated "If you want a surface to be flat, most of us would specify the flatness control. But you can also plug in profile of a surface and it would mean the same thing."

I know what you are trying to say but your example may not be the best. Flatness applies to a single surface while profile of a surface could apply on many surfaces that share the same plane rather than a single surface.

Just my 2 cents worth.



Dave D.

http://www.qmsi.ca