Converting +/- tolerance to a profile tolerance.

[Jacob Cheverie]

I have attached a basic image of the situation that I am faced with. There is a sphere and a swept (toroidal) radius. The distance from the sphere center to the radius center has a +/- tolerance. The size of the radius has a +/- tolerance. I'd like to convert this to a profile tolerance, if possible. I have tried to analyze the worst case situation on either side and my solution is to make the center to center distance basic at nominal and make the profile tolerance twice that of the tolerance on the radius size. I would use the sphere as a secondary datum and the right end cylinder as primary datum.

Any suggestions?

 

[Burunduk]

Since the toleranced location dimension (a+/-b) is from the sphere center, I would make the sphere the primary datum to constrain translational DOF and the cylinder secondary to constrain the 2 applicable rotational DOF.

EDIT; I would consider a composite profile tolerance to reflect the float allowed by the locating dimension. The second segment datumless profile tolerance value would be 2c and to reflect the float allowed by the tolerance on the locating dimension a, I would consider allowing datum shift by referencing the spherical datum feature at MMB.
The caveat is that you would not always get the entire location tolerance as a float (shift), it would depend on the UAME size of the actual as produced sphere. This is the best I can think of right now.

 

[Jacob Cheverie]

Burunduk,

I think that is a good suggestion. With the second tolerance zone being of width 2c, I must determine the size of the first tolerance zone which will be related to the spherical datum feature in an attempt to control the position.

Wouldn't the profile control simply be controlling the center point of the radius if using the sphere as a primary datum? If so, wouldn't the zone size simply be 2b?

Would this method only apply for b > c?

 

[Burunduk]

Jacob,
Yes, with the first segment of the composite value being 2b, and referencing |A|B| in that segment, it will locate the entire radius feature in the right-left direction within "+/-b". Note that it will also locate it in the "up/down" and "into/out of the page" direction within the same amount (controlling the allowed centering variation). By the way, if you wish to control the orientation of the radius feature somewhat tighter (to avoid it appearing "slanted" relative to the axis of the cylinder) you could reference B in the second segment.

 

[Burunduk]

"Would this method only apply for b > c?"
Yes. I'm afraid so.

 

[drawoh]

Jacob Cheverie,

Given that the [±][ ]tolerances shown are unmeasurable, profile tolerances are a good idea. Don't worry about the [±][ ]tolerances. Worry about what you want to accomplish, and apply measurable profile tolerances accordingly. Are you really dimensioning the centre of the spherical part from the radius?

--
JHG

 

[Jacob Cheverie]

drawoh,

That's a good way of putting it. Unfortunately it isn't my design and I was asked to inspect, so I am considering an alternative proposal without knowing what is to be accomplished entirely (other than the +/- implications). The center of the radius is in fact being dimensioned from the partial sphere.

 

[drawoh]

Jacob Cheverie,

If you are inspecting, you are stick with [±][ ]tolerances. Profile tolerances do not mean exactly the same thing.

--
JHG

 

[Burunduk]

drawoh, you said yourself in your previous post in this thread that the +/- tolerances are unmeasurable. Although the OP is at the inspection side, he does the right thing by trying to propose an alternative definition involving profile. Even if it won't mean exactly the same thing, there is no point in sticking with ambiguity.

 

[drawoh]

Burunduk,

If the drawing were under the control of the OP, I would strongly recommend redoing the drawing. It is a good request for his customer. I got into GD&T because I ran into drafting situations where it was impossible the define the part with [±][ ]dimensions.

--
JHG

 

[Vindicit]

I would use the +/- tolerances given and transfer them into profile tolerances for each feature segment. Radii are very difficult to inspect for and should, when possible, use profile tolerances to show that the feature edges are within the tolerance zone and not focus on the inspected radii values.