composite & two-single feature frame position tolerance 2

[JamesDWeed]

Group,

Anyone have a simple way of explaining the difference between composite & two-single feature frame position tolerance?...arrg.

I know the basics - top frame is for the pattern, bottom frame is for the feature, but not sure when to apply one rather than the other.

Ref pages 130 - 133 ASME Y124.5M-1994

Thanks in advance.

Jim Weed

 

[dingy2]

On a composite feature control frame, the bottom section relates inside the pattern and its orientation to datum A (perpendicular) as and example. If the bottom section is referenced to datums A & B and B is a side, it means perpendicular to datum A and parallel to datum B.

On a single segment feature control frame the bottom section if it related to datums A & B as above, it would mean perpendicular to datum A (same as composite) and parallel but also DIMENSIONED from datum B.

That's it.

Most often a composite is more applicable than a single segment feature control frame.

Dave D.

http://www.qmsi.ca

 

[PaulJackson]

Single segment fcfs are capable of controlling all six degrees of freedom (translations in XY&Z) and (pitch, yaw and roll) to the degree that the datum features are capable of constraining them.

The top segment of a composite works identical to the single segments but the bottom segment can only control orientation of the pattern (pitch, yaw and roll) to the degree that the datum features are capable of constraining them.

The pattern features themselves are always constrained for all six degrees of freedom within the pattern regardless of the datum features specified. Since the lower segment is typically a refinement of the upper the pattern itself would be constrained by that smaller tolerance.

Use multiple single segment controls when both location and orientation need to be refined in the lower segment.

Use composite controls when location is permitted to vary according to the upper segment but orientation needs to be refined in the lower.

All identically referenced and modified fcfs become one pattern according to the simultaneous requirements rule unless it is stated otherwise but the lower segments of composite controls are exempt from that rule.

Paul

 

[PaulJackson]

Sorry for the redundant response Dave and I were typing in unison. Paul

 

[powerhound]

Check this tip out on the tec-ease website. It has visual aids to help those of us who are more visual learners.

http://www.tec-ease.com/tips/december-98.htm

Just a side note. This May's tip is my suggestion based on questions arising out of our QC department. Don even used the drawing I made and answered my original question that prompted the suggestion for a tip. I was all excited that he used my tip and went ahead and signed a few autographs at my shop...haha

Powerhound
Production Supervisor
Inventor 11
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[JamesDWeed]

Thanks everyone...I always learn something from this group. Composite is what I will use...the two single FCFs almost seem redundant, but i'm sure there's a case when it applies.

Jim Weed

 

[namwob]

A simple way to think of composite tolerance (not two single-segment) callouts is that the top frame PLTZ controls the hole pattern to the datums (often the edges of a part) and the lower frame (FRTZ) controls hole-to-hole (and perpendicular to A).

Also, it's interesting to note that so many people think that .014 is an approximate conversion to +/-.005 linear, but it's not quite true for hole patterns. .014 is approximately +/-.005 at EACH location/hole. So a .014/.007 composite tol would be the correct way to control a hole pattern similar to a +/-.005 from edge and hole-to-hole linear dimensioning scheme. Only using one feature control frame back to ABC is equal to baseline dimensioning scheme - that's why composite exists.

 

[ringman]

A point perhaps oftimes overlooked is that he PLTZ is more adequately or properly stated as rectangular, rather that diametric, dependant on the boundary of the part.

JamesDWeed,

Not stated, but presumably you are using Y14.5-1994. If not, I think there are some other issues.

 

[KENAT]

"dependant on the boundary of the part" or possibly the shape of the bolt pattern/item mounted by the bolt pattern.

I've seen nominally 'round' connectors mounted by a square hole pattern. In this case a round pattern locating tolerance zone should probably be round, shouldn't it?

KENAT, probably the least qualified checker you'll ever meet...

 

[WHITMIREGT]

A good way to understand the difference between composite and two single-segment position control is to look at a pipe flange with a six hole pattern perpendicular to the face of the flange datum “A” and located or oriented to a center datum axis hole, datum “B”.

Composite Position - The upper larger tolerance zone (PLTZF) needs to be perpendicular to datum plane “A “and located from datum axis “B” with basic dimensions.

Two Single-Segment – The upper larger tolerance zone (PLTZF) will be the same as composite position as stated above. The lower tolerance zones also position the holes to each other.

Composite Position - The lower smaller tolerance zone (FRTZF) needs to be perpendicular to datum plane “A” and ORIENTED to datum axis “B”

Two Single-Segment – The lower smaller tolerance zone (FRTZF) needs to be perpendicular to datum “A” and LOCATED on datum axis “B”. The lower tolerance zones also position the holes to each other.

Bottom line is, composite position controls “ORIENTATION ONLY to the datums, not location.

If the design requires the six hole pattern to be on the center of the flange hole, use two single-segment.

 

[PaulJackson]

Only using one feature control frame back to ABC is equal to baseline dimensioning scheme [/ color]

Namwob,

There isn’t and never was an equivalent conversion from “baseline” dimensioning to a “geometric position callout with a diameter symbol.” One is restricted with baseline dimensioning to a linear tolerance in the direction of the leader line from one feature to another and with the position callout to a diameter or cylindrical tolerance zone oriented and located “as able” from the specified DRF.

You are correct however that many think that there is a conversion. The whole debate “I think” was precipitated “at the onset” by those teaching GD&T that there is 63% greater area of tolerance using diameter tolerance zone rather than a square tolerance zone with round features and further confused by those seeking to reverse the modern geometric diameter tolerance zones to individual coordinate limitations for coordinate scrutiny.

Thoughts aside… There is an ongoing debate about whether a geometric position tolerance callout “applied to a round feature” that is not preceded by a diameter symbol invokes a “square” shaped tolerance zone …or… whether the zone remains a diameter value simply because it is applied to a round feature. I grew weary of following the debate because of its length and diversity but I think that simple baseline dimensioning does not infer that the tolerance zone is necessarily uniform “with right angles” simply because it is drawn so…since no DRF has been established. If indeed a DRF is established by a general note or by a geometric callout and the appropriate “degrees-of-freedom” constrain the zone’s parameters relative to the DRF then it can be determined square, rectangle, circular “if preceded by a diameter symbol”, a circular segment “as illustrated in figure 5-42”, conical “as illustrated in figure 5-40”, or whatever shape. Whether it defaults to a circular because it is applied to a round feature will continue to be argued until the purists and practitioners are assuaged by a statement in the standard.

At first glance I could see some logic in your comments with respect to composite tolerances. Since the lower segment of a composite constrains only orientations “not translations” relative to the DRF. One could imagine that a features axis or pattern of features axes, center-plane(s), or point(s) ”if spheres”… could occupy the equivalent square portion of a circular boundary tolerance since there are no translation requirements for the lower segment…however…as I was writing here I withdraw. The upper controls location to the DRF… not the lower so there is no boundary for location. The lower controls the pattern’s relative location among its members and only its orientations to the DRF. There is no location requirement of the lower portion of a composite to the DRF!

JamesDWeed,

I never responded to your final comment “Composite is what I will use...the two single FCFs almost seem redundant” because I didn’t want to appear “anal” as my kids used to chide me but I think that specifying a composite tolerance would be functionally more rare than specifying two single segments since the lower only controls orientations to the DRF. Never-the-less function should be the driver.

Namwob,

The reason that… “that's why composite exists” …”I think” is because the Y14.5 committee had to make a distinction between multiple single segment callouts and the composite callout in the 1994 standard. It was not distinguished so in the 1982 standard and therefore had no difference.
I imagine that a significant amount of “spatial brain energy” and some “12th juror opinions” brought us the current composite position tolerance definition. Never-the-less I have found it useful to disassociate one pattern from another overriding the rule for simultaneous requirements which “I think” is flawed…that’s another doney-brook.

Paul F. Jackson

 

[ringman]

Paul,

As I understand your last thread, I agree 100 percent. By chance did you attend any of the Y14.5 meetings preceeding the release of 1994 Std?

 

[PaulJackson]

Ringman, No I did not.

 

[axym]

Gary,

I'm confused by the pipe flange example that you described.

If the center datum hole B is referenced as a secondary datum feature, it cannot constrain any rotational degrees of freedom. The primary datum plane A constrains two rotations, and B is not capable of constraining the third "clocking" rotation. All B can do is constrain two translations.

Because of that, I don't see how the example illustrates the distinction between composite and single segment position controls. Referencing B in the lower segment of a composite FCF would have the same effect as not referencing it.

Am I missing something?

Evan Janeshewski

Axymetrix Quality Engineering Inc.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[WHITMIREGT]

evan: I was thinking that the rotation of the flange was not important. If I need the six holes centered to the center hole datum "B", I need to use the two single-segment control not the composit position control because it only controls orientation to the datums. When this change to composite position was submitted for the 1994 standard. I gave a talk to the Y14 committee why orientation only was not a good change. However, the change was made anyways.

 

[axym]

Gary,

I'm sorry to keep harping on this, but I must. I can see that you're not a big fan of the way composite position is defined in Y14.5, and I agree that composite position does not add anything useful at all in your pipe flange example. It really doesn't apply. That's why I don't think the example illustrates the distinction between composite and single segment.

I went back to your earlier post and found the statement that I'm having a problem with:

Composite Position - The lower smaller tolerance zone (FRTZF) needs to be perpendicular to datum plane “A” and ORIENTED to datum axis “B”

If the tolerance zones are perpendicular to datum plane "A", they can't also be oriented to datum axis "B". Both of the rotational degrees of freedom that "B" is capable of controlling have already been constrained by "A". All "B" can do is locate.

The datum features referenced in the lower tier(s) of a composite FCF are allowed to control rotational degrees of freedom only, not location. So in the pipe flange example, referencing "B" in the lower tier of a composite FCF makes "B" do absolutely nothing. The effect would be no different than referencing only "A" in the lower tier. That's why I think the example obscures what composite position actually does.

A better example is if "B" is a planar surface perpendicular to plane "A". In that case, there is a definite distinction between referencing "B" in the lower tier of a composite vs single-segment FCF's. With single-segment, "B" would locate the pattern as well as orient it. With composite, "B" would only orient it.



Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[WHITMIREGT]

Evan: I agree with you. Thats the problem with composite position is orientation only not location. If you want location to datum "B" we need to use two single-segment.
I use this example to illustrate the difference between the two.

 

[PaulJackson]

I understand your point Evan and your advice Gary and I don't want to divert the discussion but...I have been thinking and working on this reply this afternoon which started out with composite tolerancing but morphed.

When Al Newmann presented the new interpretation of Composite Tolerancing at the (International Forum on Dimensional Tolerancing and Metrology, ASME 1993) I was disturbed! Not because of the new definition but because they changed the rules about what I understood to be “the meaning.” Immediately I felt “what is the possible use for that?” You can see that a large portion of section 5 of the 1994 standard is devoted to explaining the new definition. Eventually I found some uses for it as eluded earlier.

What disturbed me the most though, is how they chose to define concentricity. Prior to 1994 I had concentricity nailed! Even though the standard cautioned that the “center or central tendency of a feature was difficult to determine” I knew that that was all that a CMM algorithm for least squared fit of a circular trend line did! The definition fit perfectly with the inspection procedure and likewise captured the function for rotating components that were sensitive to balance.

I always felt that position callouts should pertain to surfaces…cylindrical, slab-like, conical, etc. instead of axes since axes do not physically exist…they must be estimated from the surface. There is a little note at the end of paragraph 5.3.2.1 that changes the tolerance zone from constraining the axis to constraining the surface when the axis (or center-plane) does not accurately reflect the surface irregularities. It can’t! It never does! It is constructed (somehow mathematically) from the surface!!! Every attribute gauge example in the standard checks the surface against its boundaries not the axis!!!

Bill Tandler presented his GD&T ~ CMM class materials to our GD&T Technical Committee in 2001 and along with it some proposed changes that he intended to present to the Y14.5 Committee. In that proposal was a suggestion to identify exactly which of the six-degrees-of-freedom in the feature control frame each datum feature was eligible or ineligible to constrain (either by default or by exclusion). If that proposal is adopted in the next standard I can only imagine the frustration it will cause with the “fences that we have erected around our understanding” but as with our new definitions of composite tolerancing and concentricity I am sure that eventually we will all grow in our understanding of GD&T, its rules, and its application. By the way I welcome the changes that Bill proposed so long ago because they address some of the concerns/dilemmas encountered in both function and analytical coordinate metrology.

Paul F. Jackson

 

[WHITMIREGT]

The 1994 standard committee changed composite to orientation only and thats the words the committee was using at the time, they did not have "two single-segment" position control. that was the time I complained about the change and the Y14 Main committee told the Y14.5 subcommittee to think it over some more. The Y14.5 subcommittee then came up with the two single-segment control. This made me happy to accept the new composite position where the datums will be planes, not axis unless the design only wants orientation to a datum axis.

 

[axym]

Paul,

Wow, you've opened several cans of worms here. The development of ASME Y14.5M-1994 is before my time, so my experience with it is all after the fact.

I won't even get started on Concentricity - perhaps that deserves a separate thread.

I agree that the "datum features only control rotational DOF's" idea is presented in a rather cumbersome way through composite tolerancing, but I still believe that it is a useful tool. The way I look at it is a means of applying orientation tolerances to patterns of features. Since Parallelism, Perpendicularity, and Angularity can only be applied to single features, this can be quite useful.

It's interesting that you mentioned Bill Tandler and his great idea of explicit degrees-of-constraint modifiers. Like many of Bill's ideas, it is a significant improvement that requires a different way of thinking and therefore gets a lot of resistance. I do some subcontract training for Bill's company, including the Advanced GD&T for CMM Inspection course, and getting his insights firsthand has been a real privilege. Bill is the absolute master of datum reference frame construction, and most of my understanding of that field (and GD&T in general) has come from him. Bill has been a huge influence on me in many ways.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca