True position of SR

[[Cl...]]

Print ask for position of a spherical radius to the larger OD (Datum -A-)
Depth is controlled by the .379 dimension. Shouldn't this callout be diametric T/P?
Datum -A- is the only datum.

[[Aa...]]

Bingo.

5.4.2 - "Where applicable, the tolerance is preceded by the diameter or spherical diameter symbol..."

[[Br...]]

If the feature is spherical then the tolerance zone symbol should also be spherical-- "SØ"
Probably the larger problem here is that the spherical radius is not a feature of size, and therefore cannot have a position tolerance applied to it. Profile is a better option for this.

[[Ri...]]

Please sign in to view this quote.

Using "should" is suggestive not mandatory.
I would roll with the FCF, more allowance for position.

[[Cl...]]

I would make the top surface (plane) datum -A-, the larger
OD, Datum -B- & make the .379 a basic dimension. Then
callout surface profile of the SR to A&B.

[[Aa...]]

Please sign in to view this quote.

Oops! I cited the wrong section. My quote was from 3.4.1 of 2009.

More to the point,

7.4.6 - "A positional tolerance may be used to control the location of a spherical feature relative to other features of a part. See Fig. 7-35. The symbol for spherical diameter precedes the size
dimension of the feature and the positional tolerance value, to indicate a spherical tolerance zone."

Neither section uses "should".

[[Ri...]]

Nothing changed then, may or may not. (No disrespect meant)

[[Br...]]

Please sign in to view this quote.

To be fair, If you read the rest of 7.4.6 it says with a special indication you can have a different shape tolerance zone. Unfortunately there's no such indication. Are we to suppose the tolerance zone is a cube to be evaluated in 3 directions? None of this really matters anyway because you can't properly determine the center of the actual mating envelope without a feature of size. The CMM will simply come up with some mathematically projected sphere based off of what little data it gets from the actual part and give some sort of a center point based on that. Not the most useful information. If it fails, it doesn't mean the part wont function because this doesn't really reflect most design intent and the entire purpose of the position tolerance.

[[Ja...]]

"the spherical radius is not a feature of size"

Brett, I need some clarification. Why do you consider a sphere not to be a feature of size?

[[Cl...]]

https://www.tec-ease.com/gdt-tips-view.php?q=94

[[Jo...]]

I'll chime in while waiting on Brett to respond. He said "the spherical radius" referring to the sphere in the OP, not "a spherical radius" which could be interpreted to mean any sphere at all. The sphere in the OP does not have opposed points. This is why it is not considered a regular feature of size, which also means you can't use position to locate it. The sphere in the snip from the standard does have opposed points so this is why position would be appropriate in that case.

[[Br...]]

There needs to be directly opposed points. If the sphere had more the 180° of surface it could then be considered a regular feature of size. Basically, if you can pinch the feature on both sides to measure it with a caliper, its a feature of size.

Edit: Ahh John beat me too it. he explains it better anyway.

[[Ja...]]

Guys, I’m not trying to start anything here, but this sounds like this is going to be one of those “let’s agree we disagree” things. I’ll state my viewpoint and move on.

Yes, the standard mentions two opposing points, but I highly doubt they meant it to be taken so literally. I do realize it’s the standard we go by but I think some common sense needs to be applied on my part to recognize that, in this case, a spherical radius comes from a sphere. A sphere has an infinite number of opposing points, and a sphere can always have a size assigned to it. If a radius is not a feature of size how can we assign a size to it - In this case, how can we assign SR 0.748? And how else would you describe it without adding an inordinate amount of information to the print? A sphere is a sphere regardless of how much of it is present just like a circle, cylinder, cone, torus, etc. To say that a location of a radius can’t be dimensioned because of how much arc is or is not present doesn’t make practical sense to me. If that were the case we could never have partial features such as a corner radius or other similar features without having the potential of adding an awful lot of complexity to the print. If the authors did mean it literally, then I'll abide by it. But I still wouldn't agree with it.

[[Jo...]]

I'm not sure if you're looking for understanding or not since you opened with "let's agree to disagree" then you stated your case.

[[Ja...]]

Just wanted to throw my viewpoint out there. If you agree with it, great. If you don't, that's ok too. We'll probably never know the definitive answer unless one of the authors gives it to us.

[[Br...]]

Clarke Gilbert posted a link from one of the committee members--The late Don Day. This shows opposed points
https://www.tec-ease.com/gdt-tips-view.php?q=94

Here's a Q&A with Alex Krulikowski--another commitee member of Y14.5. Look at question 5 about radius' being features of size.
http://www.etinews.com/etimail/archive/ ... /#q5answer

Those guys are just a couple of the "authors" ,as you put it, that say that a regular feature of size need to have directly opposed points.

John Acosta (who posted above) also teaches Y14.5, is certified at the senior level in ASME Y14.5, regularly goes to the Y14.5 committee meetings, and knows the committee members. (He's on also on the Y14.36 committee) So he's a pretty good authority on this as well.

But instead of appealing to authority, lets unpack some of your thoughts and see if we can make sense of what Y14.5 means objectively.

Please sign in to view this quote.

First off, having a thorough understanding of Y14.5 and correctly applying it is hardly what i would consider "common sense" especially when dealing with complex geometries. So I would beware throwing that term around. A spherical radius comes from a sphere, but it is not a full sphere. Its a partial sphere. A full sphere would have directly opposed points, but this partial one does not. Think of it this way-There are no vectors that point at each other on this surface.

Please sign in to view this quote.

The "SR" a is actually not a size dimension. Its a surface call out. The symbol for radius "R" when referenced on a drawing creates a zone defined by two arcs (minimum and maximum radii). The part surface must lie within this zone. This is similar to a profile tolerance.
Features of Size like a full cylinder, depicted by the diameter "Ø" symbol, are subject to Rule #1. This is really where the radius feature of size argument falls apart. Rule 1 at its most basic terms can be defined as having perfect form at MMC. What this means is that there is a boundary or envelope at MMC that has perfect form which the feature of size cannot violate.

Let me give you an example. If we have a hole with a dimension Ø20.15±0.05, its MMC size is Ø20.10. So this hole if manufactured at Ø20.10, must have perfect form. As it departs from MMC, it can begin to have form error as long as none of the surface elements violate the boundary. This is essentially what your doing when you check a hole size with a gauge pin.
With a partial sphere, it could do all sorts of strange things and not violate the boundary. It could completely flatten out since there's no opposed side to break that boundary.

When we apply a position tolerance to a feature of size, we need the actual mating envelope to produce an axis, center plane or center point. An actual mating envelope is defined in the standard as an envelope outside the
material. A similar perfect feature(s) counterpart of smallest size that can be contracted about an external
feature(s) or largest size that can be expanded within an internal feature(s) so that it coincides with the surface(s)
at the highest points.

In our case we have a partial sphere. An actual mating envelope would expand forever as it never contacts the opposite side of the sphere. A center-point cannot be produced this way.

Please sign in to view this quote.

No one is saying that a radius cannot be dimensioned. We're only saying position is not the appropriate way to control it. Profile is a perfectly fine option for controlling the surface of the spherical radius and will best reflect the design intent.

Y14.5 already handles fillets or small corner radii. There's an implied tangentcy rule. If it looks tangent, then it is. All you have to dimension is the radius itself, and the location is controlled by the tangent edges.

[[Jo...]]

There are definitions for certain things and there are reasons for those definitions. I can assure you that the committee means for people to take the definition of a regular feature of size literally. I've been attending the Y14.5 meetings twice a year since 2014 and have sat through the discussions. Admittedly, there is even disagreement among the experts about certain things, but this particular thing isn't one of them. There might be some differences on finer details of a regular feature of size but whether or not it must have opposed points is one of the fundamental parts of the definition. A sphere does have an infinite number of opposed points but we're not talking about a complete sphere here. This is a portion of a sphere which has no opposed points at all. Not all dimensions are the same. You can put a dimension on a slot depth but it won't be a size dimension. When you put a dimension on the diameter of a pin, it is a size dimension. We can always disagree but there is an international standard that tells us how things like this are done. If we are working to that standard, I think we should follow the rules of that standard. Obviously you don't have to if you don't want to.

EDIT: I see Brett and I were posting at the same time. Great explanation Brett.