Must profile be applied to a basic profile?

[Belanger]

Several things have popped up in various threads that always seem to leave this question lingering: When a profile tolerance is applied, is it required for the dimensions defining the shape to be basic?
We all agree that the dimensions defining the location of the feature do not have to be basic (often they are). So please realize that I'm talking about the shape -- such as the radius of a curve: I would say that the radius must be basic if you're going to apply a profile tolerance to such a curve.

See the attached picture for the specific question to be debated. Would you say the two drawings mean exactly the same thing? Or would you say the second drawing does not comply with Y14.5?

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

 

[pylfrm]

Perhaps I'm jumping in on this a bit late, but I have a few thoughts on the cone figures in Y14.5-2009 and how they might relate to other shapes.

First, a conical surface in isolation has a single parameter: angle. There can be no notion of size for a cone without involving other features / surfaces / datums / etc. What might be thought of as "size" is really just a matter of what portion of the surface you're looking at.

In Fig 8-17, the true profile of the cone is fully defined by the basic angle. The profile tolerance has no datum references, so "conicity" is all it can control. The way the I see it, the diameter tolerance just controls where the flat face ends up relative to the cone. It has nothing to do with the profile tolerance. Even if it were basic, the meaning of the profile tolerance would be unchanged.

I don't think Fig 8-17 can be directly extended to the original cylindricity/profile example though. When the basic angle goes to zero and it turns into a cylinder, a second parameter is required: diameter. This was not provided as a basic dimension, so I don't think the definition of a true profile in paragraph 8.2 was met. I'm still unclear on whether a profile tolerance can be applied without a true profile though.

In Fig. 8-18, the cone is not fully related to datum B by basic dimensions. Orientation is already controlled by datum A, so I don't think datum B actually changes anything for the profile tolerance (if you accept the idea in Fig 8-27). The "Means this" explanation implies this as well: when describing the profile tolerance boundaries, datum B is not mentioned. Again, I think the diameter tolerance (this time at the gage line) just controls the relationship between the cone and flat surface, not really "size" of the cone.

Or am I off-base here?

- pylfrm

 

[Belanger]

pylfrm,
I guess I don't really have an issue with Fig. 8-17 because there are no datums. The profile tolerance can slide (or shift) left and right until it superimposes properly along any diameter. It is controlling form, but doesn't have any way to control diameter because there is no datum in the longitudinal direction. (With my original graphic of a cylinder, there was no taper to speak of so the diameter to me should have been basic.)

But Fig. 8-18 sticks in my craw because of datum B, which locks the profile tolerance at a precise left/right location. So my thought is that the diameter shouldn't be basic. But based on the discussions here, I suppose the consensus is that it's technically legal. Shrug.

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

 

[dgallup]

We have multiple geometric controls available for cylinders: runout, total runout, cylindricity, true position & concentricity. What can you possibly gain by trying to use profile to control a cylinder when you have so many other, more clearly understood controls available?

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[drawoh]

dgallup,

I tend to place a sloppy profile tolerance all around my parts. Then, I apply the other controls, particularly parallelism and perpendicularity on the features that require them. On its own, profile does control everything. Generally, you don't want to make everything ultra-accurate.

--
JHG