Need help understanding GDT call out

[cemg]

I am a materials engr and a bit rusty on my GDT skills. Can someone out there help me to understand the positional tolerance call out in the attachment? I have an idea but not 100% sure I am interpreting correctly.

 

[drawoh]

cemg,

The positional tolerance locates the pattern, then it locates the holes within the pattern.

Within the pattern, each hole must located somewhere inside a Ø1.5mm diameter. The pattern as a whole must be located such that each hole is located inside a Ø5mm diameter.

--
JHG

 

[3DDave]

Because they are separate callouts you can determine part orientation WRT datum A separately when inspecting each FCF requirement. Were they a single callout, the lower FCF would be a refinement on the threads in the same DRF.

Also, is something wrong with an M12 callout on a hole that is 1.75 deep?

 

[3DDave]

Never mind on the thread callout. I presume 1.75 is thread pitch, not thread depth. Stupid metric threads.

 

[Belanger]

To really answer this, we'd have to know what datum feature A is. But assuming the typical situation where A is a plane that the holes are to be perpendicular to, then there is no difference whether the position symbol is shown once or twice.

The axis of each hole must fall within a 5 mm dia cylinder, with respect to the three datums. But the designer still wants to keep the hole-to-hole location somewhat tight; that is handled by the lower callout's 1.5 (each axis must also be within a 1.5 dia cylinder, where that cylinder is locked with respect to the other 4 holes, and perpendicular to datum A).

Sorry if I'm rehashing what the other guys said -- just wanted to add the nuance about assuming datum A as perpendicular.

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

 

[3DDave]

True, but only if it's not a rocker.

 

[Belanger]

Not sure what you mean, Dave...

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

 

[3DDave]

If the datum feature can allow the part to rock, then there are multiple solutions to determining part orientation relative to the datum simulator.

 

[Belanger]

Yes, and that's fine. The bottom feature control frame is asking for us to only sit upon that one feature and then verify the five holes. (No matter if it's one or two position symbols.)

An analogy... think of a simple parallelism check of one face of a block to an opposite face. That too is orientation to only one datum; would that be problematic because there might be a possibility of rocking? No.

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

 

[3DDave]

What I was referring to was the original drawing can have two solutions for resolving part orientation of datum feature A to datum simulator A, one for each independent datum frame, as opposed to it being a compound FCF in which there would only be one resolved orientation for both segments.

This situation is often disguised by manufacturing methods that are better at forming parts than is necessarily toleranced.

In the example of parallelism, there is only the requirement that one orientation of the part meet the requirement. So, one side of the part could actually be spherical and meeting the requirement requires rocking the part to be close to one edge in order for the other side to meet the parallelism requirement.

There was an attempt at regulating the selection of candidate datums in 'Y14.5.1, but it's pretty rare that I see any mention of that spec, so I believe few people depend on it.

It's an interesting problem. While the 'Y14.5 standard discusses the validation of features relative to datum reference frames, it is a little weak on discussing placing parts relative to datum simulators, beyond mostly assuming that parts conveniently conform to their datum simulators. Perhaps the '2009 version is improved in this regard.

 

[Belanger]

I still don't know if I agree, but I haven't thought much about that rocker situation. I meant that even if it were a composite FCF (with one position symbol), there is not "one resolved orientation" for the two segments; an inspector can pick the part up off of the upper segment's gage and drop it onto a separate gage for the lower segment without replicating the exact seating (assuming this is a rocker).

Are you saying that on a gage for a composite FCF the two walls simulating datums B and C would have to suddenly drop away while carefully preserving the same three contact points on A for the larger gage pins? There isn't anywhere in the standard that details a difference between one position symbol vs. two symbols when there's only one datum on the lower segment, so this is a leap I am not comfortable making

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

 

[3DDave]

For composite tolerance verification, the theoretical gage would be a compound gage.

For the typical plate-with-holes example, the gage would have two components. One component could have two sets of pins, one set on each side of a plate. One set of pins engages the holes in the part to simulate the lower segment and their size would represent the tolerance allowed there. The second set of pins would mate with holes in a base-plate and the difference in hole size would be the difference remaining between the lower and upper segment allowance. The base plate would have secondary and tertiary simulators.

This inspects the upper and lower segments simultaneously with the part in one orientation.

For two single segments, there would be two separate gages with orientation of the part determined independently for each gage.

By not sharing identical DRF references, single segment tolerances don't invoke simultaneous requirements.

Now here's the funny/stupid part. In '94 5.4.1 (p95), it says of composite tolerances "Each complete horizontal segment ... may be verified separately, but the lower segment is always a subset of the upper segment."

To my way of thinking, if one reorients the part between verifications there is no guarantee that the lower segment is a subset of the upper segment. Some may just stop with the phrase 'may be verified separately' but I think it is sloppy thinking as it ignores the subset requirement.

 

[Belanger]

Interesting thoughts, but I don't think that's how it is to be interpreted. When the only datum in the lower segment has a perpendicular relationship to the holes, then there is no difference at all between showing one position symbol vs. two position symbols. You quoted it yourself, saying that "each complete horizontal segment ... may be verified separately." I don't know why you insist it must be a "compound" gage when it clearly says gage them separately!

When they say that the lower segment is always a subset of the upper segment, they mean that the tolerance number of the lower must be a refinement of the upper tolerance. They also mean that any datums referenced in the lower segment must be repeated from the upper, and must be of the same precedence. But it doesn't require you to hold the exact same three points from the upper segment's gage, if it's able to rock. (It's wise to do so, but we're discussing an academic point, I suppose.)

Your interpretation of a supposed difference between one position symbol and two in this case is a novel idea, but I don't think it really holds up. But it might be interesting to hear others chime in on this.

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

 

[Tunalover]

I think that the two separate FCFs have different meaning than a composite FCF. I think that the guy who did the drawing intended one composite tolerance FCF.

Tunalover

 

[Belanger]

In the case of datum A being perpendicular, there is no difference. But yes, it's more straightforward to have used one symbol.

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

 

[cemg]

thank you for your help everyone...I think I understand
and yes, datum A is a plane that the holes are in.