Profile tolerance for toleranced feature

[Gam]

Hi all,
Could you give a look at the attached file and let me know if this callout is legal?

It is a profile tolerance of .1 with no datums, applied to a surface of a sphere with a toleranced radius.
I think it is ok but I need to convince other people.
If so, could you also point me to the relevant section (or example) in the ASME std?

Thanks

Gam

[Kelly_Bramble]

The radius symbol on the limits of size specification should be changed from R to SR. SR stands for Spherical Radius. Also, the limits od size specification is not illustrated to any know standard

The Profile of Surface specification is legal except the tolerance boundaries will allow a surface variation equal to that allowed by the Spherical Radius Limits of size (+/- .5). Therefore, if you're going to use the GD&T Profile of Surface specification to control surface form variations the tolerance should be less than total variation allowed by the limits of size.

Assuming you're using ASME Y14.5-2009 or earlier USA GD&T standard.

[Gam]

Yes, the SR was an oversight.
The rad tolerance is +/- .5, the profile .1, isn't it ok so?

[Kelly_Bramble]

The rad tolerance is +/- .5, the profile .1, isn't it ok so?

You're right, I thought I saw +/- .05, sorry.

[Belanger]

I would say that the spherical radius itself should not have a tolerance at all. Rather, it should be a basic dimension (a boxed dim, sometimes called a theoretically exact dimension). This is because profile tolerances must be applied to a "true profile" according to paragraph 8.2 of the ASME Y14.5 standard. And a true profile is one where the actual curvature of the surface is defined by basic dimensions.

[Kelly_Bramble]

I would say that the spherical radius itself should not have a tolerance at all. Rather, it should be a basic dimension (a boxed dim, sometimes called a theoretically exact dimension). This is because profile tolerances must be applied to a "true profile" according to paragraph 8.2 of the ASME Y14.5 standard. And a true profile is one where the actual curvature of the surface is defined by basic dimensions.

Reading further into paragraph 8.2 -> "Where used as a refinement of a size tolerance created by toleranced dimensions, the profile tolerance must be contained within the size limits."

[Belanger]

I think that statement makes sense when referring to dimensions that tie the surface in question back to a datum. (So in Fig. 8-27, the 0.07 is a refinement of the ±0.2.)

But we're talking about a case where the very curvature of the feature in question is being toleranced with profile. I wouldn't do something where the very curvature is defined with toleranced dims and then apply a profile tolerance.

For instance, could I take the classic example of cylindricity and change the symbol to profile of a surface? I would say no, unless the diameter were changed to a basic dim. That would be the closest analogy to what the OP was asking (I think).

[Kelly_Bramble]

For instance, could I take the classic example of cylindricity and change the symbol to profile of a surface?

Why not? I do not see any limitations in the standard preventing this specification substitution. For that matter one could substitute Profile of Surface for Flatness, Profile of Line for Straightness. If applicable (big IF here) one could use Profile of Surface and Line to control all features on a part or assembly.

For the record, I don't encourage this application of Profile tolerances but rather embrace the more expected form controls (flatness, straightness, Cylindricity, Circularity).