Runout with MMC/MMB on DRF? No. Profile with MMC/MMB on DRF? Yes

[greenimi]

What could be the reason that for profile tolerance, MMC/MMB modifier on the DRF (datum shift) is allowed per the standard, but for the runout using MMC/MMB in the same place (DRF-datum shift) is illegal (why the tolerance zone cannot be moved, translated/rotated with runout callout). Both GD and T tools we are checking the surface.
Thank you

 

[CheckerHater]

All other conditions being the same, profile will control surface wrt surface.
Runout will control surface wrt axis. Apples and oranges?

 

[Belanger]

Well, a surface is never a datum; a datum is always an axis (for what we're talking about here; other choices are plane or line/point). I think the reason has to do with the nature and intent of runout. It is meant to control wobble, and that is inherently a "regardless" thing.
Others on this forum have pointed out that runout is a little unusual because it is defined in the standard based on its inspection method.

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

http://www.gdtseminars.com

 

[CheckerHater]

Once you say MMC/MMB you are dealing with virtual condition of some sort of a boundary to establish your datum.
In 2009 world you can define profile wrt “complex shape” which doesn’t have an axis at all (at least position of axis is arbitrary).

Runout requires by definition axis derived RFS to begin with. This if what I meant.

We are talking two completely different tools used to achieve two completely different goals.
Is “why screwdriver is different from wrench?” legitimate question, or just troll?

And I simply cannot help but notice that runout is not unique; cylindricity / roundness also comes with inspection method attached; just because it is hidden in separate book, doesn’t make it less true. Also I would avoid such patronizing term as “wobbling”; depending on the design intent runout may be legitimate replacement for concentricity and position. In that sence runout is no less powerfull than profile.

Of course, if grandma had mustache, she’d be grandpa. If you somehow defined runout at MMB it would be not much different from profile.

 

[Belanger]

CH -- I'm not sure what you mean by "profile wrt complex shape" -- profile may be related to one or more datums, but a datum is always a "point, axis, line, plane, or combination thereof."

Profile itself can be applied to a complex shape, but that's getting to the point I was making: that runout has a more focused purpose than profile, which means that it's going to have more focused rules such as the RMB requirement.

And I don't see why wobble is a patronizing term; I was merely trying to use layman's terminology to muse about the differences between runout and other geometric characteristic controls. Certainly runout is more involved than such a casual term, but that's a perfectly valid way to explain the root intent of where runout came from. Profile is intended to be a more broad concept, so it doesn't need to have such a stringent rule about using modifiers on the datums.

The bottom line is that runout can't use MMB because that's the way they wrote the rulebook


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

http://www.gdtseminars.com

 

[CheckerHater]

JP,
From the top of my head I cannot find legitimate use for axis derived at MMB.
Is it even legal? Is there a good example somewhere?
I am not trying to pick a fight, just curious.
For example, any CAD data has 0 point and XYZ, so you always can derive point and three axis from any “mathematically defined” surface, no matter how ugly. But is there any good use for axis that shifts?

 

[CheckerHater]

Sorry, I probably made it even more confusing.
I understand datum shift when controlling position of the bunch of holes.
Since OP was about runout, I am still thinking about controlling some sort of coaxiality / eccentricity condition.
What do we gain from controlling one axis in relation to another axis that is not really fixed in space?

 

[Belanger]

A simple example would be a counterbore with a position tolerance to the thru-hole that it sits upon. The axis of the thru-hole can be derived on an MMB basis; the counterbore could then wander a little further off center from the actual thru-hole.

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

http://www.gdtseminars.com

 

[fsincox]

It is like asking why 1+1=2, it is defined that way, part of the its basic definition!

CH,
I think the same about concentricity, why do I not want to have the option of controlling an axis to an actual axis and not just an AME of a feature? I guess I should be German they seem to understand me.

 

[CheckerHater]

Yes, understood.
There isn’t much difference between two holes on top of each other and two holes side-by-side.
You probably wouldn’t apply runout to every counter-bored hole anyway
My problem is over-thinking. Too much of “devil’s advocate” play.
If somebody wants runout defined at MMB maybe there is some use for it, and we could learn something here... NOT.