Composite Position with separate requirement for pairs of features

[Huddman]

Attached is a view of a base plate that will hold three brackets, each with 2 clearance holes for #8 Screws. I have a composite position control. I want to loosely get the pattern located within .125 cylindrical tolerance zones. I then need the Y direction spacing to be at the .750 basic hole to hole position with a tolerance of .014 diameter for the cylindrical zones, but only in the Y direction. I still want the individual pairs of holes to be able to float left to right, as 3 separate units of two holes and not as a pattern of 6 holes, within the .125 pattern locating tolerance zones. I also want the pairs of hole to have the .014 dia tolerance zones perpendicular to datum B so when the assembly is mounted on the wall, the brackets are not rotated too much about the Z axis. How do I convey that with what I have got so far, per the ASME Y14.5 2018 standard? Thanks in advance.

 

[3DDave]

What you are asking for is not a composite tolerance.

Use multiple single segment tolerances, one for each component of the position control you want and use "SEP REQ" as necessary to separate the various effects as required. Look for "Bidirectional Positional Tolerancing, Rectangular Coordinate Method"

 

[Huddman]

Thanks for the reply. I thought of using SEP REQ, but I did not know how to keep the pairs lined up vertically. If I use sep req with the bi-directional rectangular method, couldn't one of the paired holes move left and one move right, allowing the bracket to be out of perpendicular orientation with datum B? I also do not want to be locked in to and basic dimension from the datums on the lower segment so I did not use multiple single segment position. I just want to maintain the .750 spacing of a pair of holes and the perpendicular aspect to A and B. That pair of holes can freely float anywhere in the .125 Dia Pattern locating tol. zones. It's like I need to position a pattern of 2 holes 3 places separately, can this be done? I'm going for the asthetic look rather than accurate position.

 

[Burunduk]

The problem with the suggestion of multiple single segments with SEP REQT at the second segments - it would also make B control location in "Y" (normal to B) within the smaller zones for each pair of holes. You seem to want just the spacing controlled in "Y" which is not the same. I suggest using 3 separate composite feature control frames for the 3 pairs, with the same datum referencing as already in your picture. The top segments would be a simultaneous requirement. The second segements will not form a simultaneous requirement (unless modified SIM REQT), because of the Y14.5 rule that defines so. So each second segment would control spacing and the orientation including rotation about "Z", but they would do so separately by default.

Datum feature A is not defined.

 

[3DDave]

There's no problem with it. You already know why, because I told the OP exactly how:

"Bidirectional Positional Tolerancing, Rectangular Coordinate Method"

What the bad triple composite method does is hide the desired tolerance zones by making them appear to be diametral when they are not. What a bad plan that is.

Clearly there is a typo that Datum Feature A has been mislabeled as "D". The better nit-pick is that Datum Feature C should be perpendicular to both "A" and "B".

 

[Burunduk]

Had the purpose been controling location relative to the datum reference frame looser in one direction and tighter in another direction, then Bidirectional would be the right choice. But since the purpose as stated is "I want to loosely get the pattern located within .125 cylindrical tolerance zones. I then need the Y direction spacing to be at the .750 basic hole to hole position with a tolerance of .014 diameter for the cylindrical zones, but only in the Y direction." The 3 composite tolerances will do exactly as required, including the orientation constraints to take care of "the pairs of hole to have the .014 dia tolerance zones perpendicular to datum B so when the assembly is mounted on the wall, the brackets are not rotated too much about the Z axis."

 

[3DDave]

My solution provides an explicit workable answer, one that is explicit about the zone and doesn't require your usual head-scratching puzzles for QA and other engineers to work through.

You cannot have diameter zones perpendicular to [A|B]. That's a width and needs a tolerance that controls ONLY the width, not the diameter, of the tolerance zone.

Using "diameter" when there is no "diameter" being controlled is misleading. Is it wrong to be misleading?

I think it is wrong to be misleading, but perhaps you disagree.

 

[pmarc]

I would propose the following solution:

1. Show XYZ coordinate system in at least two views of the drawing.

2. Control each pair of holes with two separate callouts:

2a. First callout - single segment position tolerance of dia. .125 relative to |A|B|C|.

2b. Second callout - single directional (in Y direction) position tolerance of .014 width relative to |A [u,v,z] | B [w] | with additional SEP REQT notation added to it.

With this done, the directional position tolerance does not control:
- location of the holes relative to B within .014.
- spacing between the holes in X direction within .014,
which, per mu understanding of the design intent, need not to be controlled that tightly.

 

[Burunduk]

3DDave,
The requirement was "zones perpendicular to datum B so when the assembly is mounted on the wall, the brackets are not rotated too much about the Z axis." The intent is clear; he wants each two zones kept 'vertical' (from top view perspective) which will be achieved by the rotation constraint applied by the secondary datum B reference in the lower composite segments. Tightening the location in one direction by a bidirectional tolerance doesn't seem to be related to the OP's intent.

 

[3DDave]

Any questions about adding geometric characteristic controls to the stock material?

 

[Burunduk]

I would propose the following solution:

1. Show XYZ coordinate system in at least two views of the drawing.

2. Control each pair of holes with two separate callouts:

2a. First callout - single segment position tolerance of dia. .125 relative to |A|B|C|.

2b. Second callout - single directional (in Y direction) position tolerance of .014 width relative to |A [u,v,z] | B [w] | with additional SEP REQT notation added to it.

With this done, the directional position tolerance does not control:
- location of the holes relative to B within .014.
- spacing between the holes in X direction within .014,
which, per mu understanding of the design intent, need not to be controlled that tightly.

pmarc,
Why single directional?
That would result in 2-parallel-planes tolerance zones along the X direction. Then one hole in a pair can be dislocated as far to the right and the other as far to the left as the larger zone of .125 allows. I don't think it's a good thing considering the intent to tighten the spacing within each pair for the brackets to fit.

 

[pmarc]

Burunduk,
This is how I understand the OP's description. In particular this:

"I then need the Y direction spacing to be at the .750 basic hole to hole position with a tolerance of .014 diameter for the cylindrical zones, but only in the Y direction".

Whether this is valid functional need or not, is a separate conversation to have.

 

[Burunduk]

pmarc,
I suspect it's just not very accurate way to say he wants each pair of 2 holes aligned accurately along Y (perpendicular to datum plane B).
See also in post #3: "If I use sep req with the bi-directional rectangular method, couldn't one of the paired holes move left and one move right, allowing the bracket to be out of perpendicular orientation with datum B?"
So I guess the tighter tolerance zone needs to limit X direction (parallrel to datum plane B) opposite translations.
But I guess it's up to the OP to clarify.

 

[pmarc]

Yes, it would be good to hear clarification from OP, but you may be right about the intent.

 

[3DDave]

Burunduk, so close - the vertical Y spacing is precise, the horizontal X location can vary as long as they pairs of holes shift together and retain a small, mutual, perpendicularity width relative to B.

 

[fsincox]

I think you want cylindrical zones for the final (2) hole pattern for the mating holes in a bracket relative to datum "A", or "D"? (#8 Clearance holes for mounting to brackets)
Frank

 

[3DDave]

I think you want cylindrical zones for the final (2) hole pattern for the mating holes in a bracket relative to datum "A", or "D"? (#8 Clearance holes for mounting to brackets)
Frank

That would have been my second single segment controlling the position of the pairs of diameters, 3X, to [A] SEP REQT. Then one single segment controlling the position of the zone width, 3X, to [A|B] SEP REQT to ensure they weren't rotated. This on top of the large position tolerance of the holes, 6X.

This allows separately controlling the independent factors and makes clear that there are 3 independent controls instead of three composite feature control frames of three levels and having to make sure they remain coordinated if any of the tolerances is changed, and the implication to an initial review that there are 9 separate controls in play.

---

Using the custom datum reference frame is also a goodish option, though it makes the FCFs lengthy; also while "u" and "v" are used to control angles in the FCF are used elsewhere in computer graphics and other previously establish math areas as being stand-ins for unit vectors in the mapped space on sometimes complexly curved surfaces, such as adapting a rectangular bitmap to wrap around a cylinder or a torus. Everyone else tends to use Greek letters for angles and computer software writers have access to Greek letter fonts/typefaces/glyphs.

For example, angles are typically denoted as α, β, γ, or ψ, θ, φ

Look for UV mapping for what the rest of the world is doing.