Questionable true position callout 2

[End_User]

Hello all.
First time post!

I have simple circular part with a 6 hole bolt circle pattern and a T.P. calout.
The confusion arises from the true position callout being connected to a width dimension and not a hole/diameter.

First, there are only three slots not six.
Secondly, A width has no inherent position nor does it imply the use a diametrical tolerance zone.

I'm generally confused as to what the designers intent was here and I know from previous experience that getting a timely answer from this customer is not possible.
So I guess a best collective guess from my peers seems like the next best option.

ideas, thoughts, interpretations, humorous observations all welcome!

 

[axym]

Burunduk and All,

Y14.5 doesn't currently support position RFS for curved slots. There is no actual mating envelope defined, and hence no "resolved geometry" to be contained within a tolerance zone. So if the intent is to control the radial location of the slot "centerline", there really isn't a direct way to do that. There are only tools to control the surface.

Maybe it's time for an extension of principle. What we need here is to treat the curved slot as a set of infinitely short parallel-plane slots, and apply position RFS on each one. Perhaps this could be done by adding the EACH ELEMENT modifier to the position tolerance, like they did with perpendicularity in Fig. 9-16 of Y14.5-2018 (see the thread on "datum line"). That should control the centerline of each element within a parallel-line position tolerance zone, which would be in a cutting plane normal to the basic surface.

Y14.5 doesn't show an example of this, but it also doesn't say that you can't ;^).

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[chez311]

Evan,

Thanks for confirming my suspicions. As I said the more I looked at it the more it seemed like RFS could not be applied to a curved slot in that manner. Very interesting idea with the EACH ELEMENT modifier. How would you suggest controlling the relative position/clocking of the end-radii in this case with RFS? That still seems troublesome to me, for both curved and straight slots, unless they were treated as individual FOS (semi-circles).

 

[axym]

chez311,

I agree that the end radii would be troublesome if you wanted to use RFS and control the "axis" of each end. The current definition of axis as the center geometry of the UAME doesn't work with non-opposed FOS's, because the UAME is not properly contained. I think what would be needed is a different type of fitted cylindrical counterpart, that is defined in terms of maximum contact or minimum deviation instead of being the largest size. Then the axis of that counterpart could be controlled - this is kind of what happens already if the feature was measured on a CMM.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Burunduk]

Maybe it's time for an extension of principle. What we need here is to treat the curved slot as a set of infinitely short parallel-plane slots, and apply position RFS on each one.

Perhaps another way to approach this is to break it down to opposed line elements (along the depth dimension of the slot)? The note would be "EACH LINE ELEMENT"*. Maybe it would be an easier intent to communicate to manufacturing and inspection than a "set of infinitely short parallel-plane slots", which is more... ?

* re-thinking it, a better note would be "EACH PAIR OF OPPOSED LINE ELEMENTS".

 

[mkcski]

axym:

Thanks for your insight. I do not deal with slot at all and this is a good example to dissect. Is there an irregular-feature-of-size consideration in-here somewhere? Just thinking...

Certified Sr. GD&T Professional

 

[chez311]

The current definition of axis as the center geometry of the UAME doesn't work with non-opposed FOS's, because the UAME is not properly contained.

Evan,

The end-radii should have at least one set of 180 degree opposed points, I would think they could contain a UAME - right? Also not 100% sure I can visualize your "cylindrical fitted counterpart" example with maximum contact/minimum deviation.

 

[mkcski]

No one has mentioned profile-of-a-surface for this application. Any thoughts?

Certified Sr. GD&T Professional

 

[chez311]

mkcski,

I would say profile of a surface applied all-around the slot or the boundary concept applied with position at MMC would be the best solutions for a curved slot. My questions to Evan about RFS are more academic, I am hard pressed to think of a situation where such a feature would benefit from being called out with position at RFS.

 

[mkcski]

chez311:

I understand. To me the world runs on interchangeable parts - holes at MMC - so it is "king" at 85-90% of applications - my experience. RFS and LMC are the outliners at 10-15%. To be sure, some industries use RFS or LMC a lot. So, to me, studying RFS/LMC is good for comparison and a deeper/wider understanding of GD&T concepts.

Certified Sr. GD&T Professional

 

[axym]

Burunduk,

I agree that EACH PAIR OF OPPOSED LINE ELEMENTS is a better description. They're kind of 2-dimensional slot elements. There's probably no easy way to describe this, but I wouldn't use the term "set of infinitely short parallel-plane slots" in documents for manufacturing and inspection - that would be too confusing. Perhaps "infinite set of infinitesimally small radially sectioned parallel-line slot elements" would be better ;^).

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Burunduk]

Evan,
How about an extension of a principal that would produce a different type of derived center - not an axis, not a center plane, but a "median curve", and let the position control limit it within a segment of a ring- shaped tolerance zone?