Position Refinement 1

[jimbod20]

I have a question regarding position control/refinement. I provide a sketch.

D, E, F are cast datums. H, A and B are machined datums.

I locate machine surface H with a basic dimension and profile tolerance from cast datum D. I locate machine surface A with a basic dimension and profile tolerance from cast datum F.

I now want to locate machine diameter B with position to H, A and the cast datum E (cast datum E is secondary and consists of two .250 inch diameter buttons). I hold a position of .050 H, A and the cast datum E. The part is fully constrained by datums H, A, E. I need tighter control of position to machined datums H and A so I 'refine?' the position of B to H and A with a position of .015. Is this specific feature control frame correct if the part is not fully constrained by a tertiary datum? I add an additional 'refinement?' of perpendicularity of .0005 to surface A. Is my interpretation of refinement correct? Is this composite feature control frame consistent with GD&T language?

 

[jimbod20]

I provide a sketch.

Thanks.

 

[Medmaker]

Not sure from your drawing how datum E works, but I'm assuming in constrains translation in and out of the page.

What you show in the sketch is a multiple single segment positioning tolerance. In this case, each segment controls both position and orientation of the tolerance zone (see 2009 paragraph 7.5.2). It would be incorrect to repeat the same datum structure in both segments as whichever one was the tightest would override the other. What you specify in your drawing is that the bore is tighter with respect to location to H and orientation to A but is allowed to more freely move in and out of the page to the tolerance described in the first segment.

Typically, this type of tolerance, along with composite tolerance, is used with patterns of features. I'm not sure if anyone else here has used it with a single feature.

Perhaps you'd be better off using Bidirectional Positioning Tolerance as described in Fig 7-28 of the 2009 standard.

 

[jimbod20]

You are correct in your assumption regarding datum E. I did not illustrate or denote in my original post. E is the secondary cast datum beyond the surface of the page/sketch. D, E, F are primary/secondary/tertiary cast datums.

What you describe is what I want. I want the bore diameter location within .015 to H and orientation within .0005 to A yet allow the bore diameter to move more freely in and out of the plane of the page with respect to cast datum E.

I have seen a similar format on customer drawings.

 

[pmarc]

Medmaker,
I have a question to you: if I understood you right, are you saying that for a location control of a single hole/pin two single segment positional callouts (e.g. |pos.|dia.050|H|A|E| & |pos.|dia.015|H|A|) would be incorrect?
Please ignore my question if I misinterpreted your statement. Thanks.

 

[Medmaker]

Pmarc,

I do think that it is incorrect. Especially with the use of the cylindrical tolerance zones. I interpret this bore lying in a rectangular tolerance zone .050 long and .015 wide. Seems as if it would be better represented with a view down the axis of the hole with bidirectional positional tolerancing referencing all three datums.

From what it appears in the standard, all of the uses of composite or single segment positioning falls under the sections controlling patterns. Is there support for using it to control a single feature? If so, I'd love to see an example.

 

[pmarc]

Medmaker,
Agree on composite - this concept can only be used with patterns. Lower segment, by definition, merely controls relationship between features in the pattern, so it would be meaningless if there was only one feature controlled.

However multiple single segment positional controls refering to the same datums (I mean for instance |A|B|C| in the top segment and |A|B| in bottom) are not reserved exclusively for group of features. It is true that standard does not show a single example similar to what I described in my previous post, but I think you agree that valid geometrical interpretation of the concept is quite easy to figure out and that it is not in conflict with any Y14.5's rule of higher order.

And when you imagine that, you will see that the interpretation is different to bi-directional positional tolerancing especially in terms of shape of tolerance zone.

 

[Medmaker]

pmarc,

I agree with what you are saying. I've attached a drawing of my interpretation of what the tolerance zone would look like for a single segment positional control of a single feature (the cross-hatched area being the zone). Essentially it's a racetrack whereas a bilateral positional tolerance would be a true rectangle. Would you agree?

I also think we both agree that you can't fully constrain both segments with the same datum structure (in response to Jimbod20's orginal question).

 

[pmarc]

Medmaker,
Allow me first to answer to your second question. Yes, both single segments positional callouts cannot contain exactly the same datum references.

As for shape of the tolerance zone, I think it is different to what you show. Please look to attached graphic.

http://files.engineering.com/getfil...616c92a3&file=mult_singl_seg_pos_tol_zone.pdf

In general the single axis must reside within a cylinder of smaller diameter which is able to float within larger tolerance zone staying fixed at basic distance from datum B. The resultant area where the axes of different holes may lie is shown in bottom right picture, but the biggest area that the single axis can reside in cannot be larger than the smaller cylinder.

 

[leeekim]

pmarc,

I'm getting into GD&T. I'm wondering if you can clarify a question regarding the standard for patterns. In your attachement, can the axis of the FRTZF be outside the PLTZF? I know the standard says the axis of the feature must go through both tolerance zones. It appears to me this can still happen although the intersection of both tolerance zones decreases somewhat.

 

[Belanger]

I hope pmarc doesn't mind if I jump in on that question: The FRTZF doesn't have to be entirely inside the PLTZF. If some of it is outside, that simply means that there is less area for the real hole's axis to fall within. Of course this means that there must at least be some overlap -- some small portion of the FRTZF within the PLTZF.

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

http://www.gdtseminars.com

 

[pmarc]

leeekim,
To have everything clear, my example was not about pattern of features. I wanted to show how tolerance zones work when single feature is controlled by two single segment positional callouts.

For composite positional tolerancing axes of the pattern features have to simultaneously lie inside both tolerance zone frameworks, even if some portions of FRTZF lie outside PLTZF.

 

[leeekim]

Thanks for that clarification guys!

 

[Medmaker]

pmarc,

Thanks again for the clarification. I did indeed forget the concept that the small tolerance zone didn't need to entirely fall entirely in the large tolerance zone, just as long as the axis did. I think I was forcing myself to believe that the shape of the tolerance zone was going to be significantly different that rectangular. However, I do see how this shape would conform better to functional requirements.

I did lose you there though on the last sentence of your post, "The resultant area where the axes of different holes may lie is shown in bottom right picture, but the biggest area that the single axis can reside in cannot be larger than the smaller cylinder." I wasn't sure what you meant by "different holes" and what area that meant.

Thanks again for taking the time to clarify.

 

[axym]

Medmaker,

I believe I know what pmarc is trying to explain with "the resultant area where the axes of different holes may lie". The "resultant area" is the green shaded area, and "different holes" just means various possible instances of as-produced holes. A particular as-produced axis could exist anywhere within the green zone, but that axis cannot span the entire left-right length of the green zone. In other words, the axis cannot be tilted by more than 0.5. Each as-produced axis must lie within a 0.5 cylinder (red zone), but that red zone is movable in the left-right direction. I hope this helps - what we need is another diagram that shows the two zones in a side view, where we could see the tilt of the as-produced axes.

pmarc,

I don't quite agree with your statements on composite FCF's. The lower segment of a composite FCF does control the relationship between the features in a pattern, but it also controls the orientation of the pattern relative to the datum reference frame. So if a composite Position FCF was applied to a single feature, the lower segment would not be meaningless. In most cases it would mean the same thing as an orientation tolerance.

I agree with your diagram showing the "resultant zone". I wouldn't necessarily use that term to label it, but I think that the shape you've drawn is correct. But I'm also going to nit-pick your description of the two zones. Technically, the smaller zone doesn't have to float within the larger zone - the smaller zone can freely translate in the left-right direction. Each FCF is an independent requirement, and we can evaluate each one independently of the other. So an as-produced axis might conform to the smaller 0.5 zone but not conform to the 1.0 zone. Or vice versa. I agree with John-Paul's statements on the two zones - there may not be complete overlap between them. The only requirement is that the axis conforms to both zones - otherwise, there is really no relationship between the two zones. They each stand on their own. In the sections on composite Position and Profile, the standard goes to great lengths to emphasize that certain parts of the zones may be unusable. I would prefer to just treat each FCF independently and say that the feature needs to conform to each one.

If we try to model the combined effects of more than one FCF, I think that things start to get confusing. When tolerance zones can translate or rotate due to datum shift effects, I am a bit hesitant to start drawing diagrams of the areas that can be swept out to form "resultant zones". When a smaller zone with open degrees of freedom is shown as a larger "resultant" zone, there is opportunity for misunderstanding.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Evan,
Thanks for your comment and I agree with most of you said except one. But allow me to explain myself step by step.

1. I felt that my wording about "resultant area" and "different holes" will create a confusion. But you got it correct.

2. It is true that two single segment callouts shall be treated separately. However my intention was just to show a zone within which the two requirements can both be met, and not one or the other. This was the direct answer to Medmaker's question, so if it mislead someone, I sincerely apologize.

3. You are absolutely correct that lower segment of composite positional callout not only controls relationship between features in a pattern, but also orientational relationship of the pattern to datum referenced in the lower segment. However I do not agree with you that composite is acceptable for single features. If lower segment deals with an orientation only, position is not a good characteristic to choose. Why not to use perpendicularity?
I think it is the same dillemma as in case of a hole/pin assigned as secondary datum feature and controlled by positional callout to datum A which is a plane perpendicular to the axis of the hole.

 

[axym]

pmarc,

I wouldn't recommend using a composite FCF for a single feature of size. I agree that using Perpendicularity (or another orientation tolerance, as applicable) would be much better. I'm just saying that the meaning would be definable (albeit confusing) if a composite FCF was used.

John-Paul,

I wonder what you would have to say about that. If the "intrinsic purpose" of a Position tolerance is to control location, I wonder what the intrinsic purpose of the lower segment of a composite Position FCF is. The standard provides a lot of debate-worthy material on this ;^).

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

If it falls under the banner of the position symbol, is should be doing the intrinsic purpose of that symbol. So I agree with ya both -- if all you are doing is perpendicular, then break it out as a separate perp symbol.

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

http://www.gdtseminars.com

 

[axym]

John-Paul,

Here's one for you. Let's say that I want to refine the orientation of the closed shape in Fig. 8-24 of '09. I can't use an Orientation tolerance because the feature isn't a planar surface or a regular feature of size. One way to do it (and possibly the only way) would be to add a lower segment to the Position tolerance and change it to a composite FCF. The lower segment could be POS|dia 0.2(M)|A|B| or something like that. The lower segment of the composite Position FCF would act purely as an orientation tolerance, refining the orientation of the feature without controlling its location.

A Position tolerance controlling orientation only. What do you think?

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Oh man, you're trying to put me right in between CheckerHater and MechNorth! (Re: extension of principles.)

I'm tempted to say that it's OK, except for the clear statement in paragraph 7.5.1 that a composite position tolerance provides "for the location of feature of size patterns as well as the interrelation...of features of size within these patterns."

Since Fig. 8-24 has no pattern, it can't hold up one of the core parts of that definition. So in a strict sense, I read the standard as not seeing composite position fit for use in Fig. 8-24.

OK, says me to myself, how would we accomplish what you ask? I dunno. One alternative might be to try to do something like profile all around wrt |A|B|, but with non-basic dims tying it back to the datums, a la FIg. 8-27. But that gets a little goofy, in my opinion, what with the basic/nonbasic dims, and it would conflict with the desired form control which is 1.2 mm. Gosh, use the T modifier on the new 0.2 profile? Ugly, man. So as I said, your proposal might end up being the best.

At any rate, this wouldn't be license to use the position symbol when perpendicularity is the only thing being achieved. (I know you weren't saying that, but one more nail in the coffin doesn't hurt.)

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

http://www.gdtseminars.com