Center of a datum feature with big tolerance

[kakalee1]

Hi,

I have a question related to center of a circular feature with big tolerance. Please refer to my attachment for better clarification. If Datum B is chosen to be a Datum feature, and it has a tolerance of +/- .020. The 2.500 circle has a true position callout for a .004 diametrical tolerance to datum B.

Here is what I am having trouble. If datum B has tolerance of +/-.02, and there is no form control on datum B. Thus, its center can be anywhere within the .04 diameter. How can I locate the center of datum B to find out if other features are within tolerance? Or maybe, my example is just plain wrong?

All feedback are greatly appreciated.

KL,

 

[Dougt115]

You can't. Try making the center the Datum instead.

 

[ewh]

There is a form control on datum B... ±.02.
The other features are located relative to the axis of B regardless of the feature size of B. The axis of B can be anywhere within the ±.02 tolerance, but the other features are located to that axis within their own tolerance zone regardless of where that axis may lie in its ±.02 tolerance zone.

“Know the rules well, so you can break them effectively.”
-Dalai Lama XIV

 

[3DDave]

You need to find the minimum diameter that the outer surface fits within while oriented to A. The center of this is the center of Datum B.

I wonder what the mating part looks like. Does it shrink down around B and locate it such that the other features are located to the outer surface, like a large drill chuck or collet?

 

[powerhound]

Doug,

That is the way you make the center of the OD a datum feature. Attaching a datum identifier to the actual center of this part is prohibited. Is that what you were trying to say?


John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[kakalee1]

@3DDave: there is one more part that will sit on top of this part, and it is located using the three .250 holes.

What I have trouble with is when I use something like a 3 jaws chuck to locate the outer surface, the center of the OD moves if I rotate the part. Imagine if I put the part on to a chuck and mark the center of the part. Then I take the part out, rotate it by 90 degree clockwise, then put it back on to the chuck. The center is now at a different position. Which centers should I use to verify that the center hole and the three .250 holes are within tolerance?

 

[JNieman]

Would it be unacceptable to make the +/-.005 diameter bore the Datum B? Then relate the +/-.020 O.D. to Datum B, rather than the other way around? Unless the O.D. is used to interface with something, that would seem more in line with design intent, and would give better control.

Otherwise, unless you have some manner of 'precision 3 jaw chuck' then you are left with using a 4 (or more) jaw chuck and centering it manually and then reading where the bore is.

I'm not terribly well versed with niche-inspection tools, but possibly there is a 3-roller-bearing setup, in which you have two bearings on the bottom set up as to hold the part akin to a v-block, and then a third atop, which can hold it down and trap it. That would allow the part to rotate freely about its center axis, while a dial indicator can check the runout of the inner circle, which would essentially yield the deviation of I.D. center-point from O.D. center. Correct me if I'm wrong, I'm kind of pulling that one from my posterior.

_________________________________________
Engineer, Precision Manufacturing Job Shop
Tool & Die, Aerospace, Defense, Medical, Agricultural, Firearms

NX8.0, Solidworks 2014, AutoCAD LT, Autocad Plant 3D 2013, Enovia DMUv5

 

[3DDave]

kakalee1,

The problem with the three jaw chuck is avoided by having an infinite-jaw chuck. Since those aren't easily available, one is left with measuring a large number of points on a CMM and then finding the smallest circle that envelopes all of them.

http://en.wikipedia.org/wiki/Smallest-circle_problem

Also

http://www.cs.uu.nl/docs/vakken/ga/slides4b.pdf

To do a useful job it is necessary to see how the part is installed and used and what is expected of it there.

 

[TheTick]

Why not add a form control to the OD. Good place for runout.

 

[CheckerHater]

kakalee1,

In the assembly is 10.00 OD actually contacting any parts?

 

[powerhound]

kakalee1,
You don't measure from the center of the part, you measure from the datum that is established by your measuring equipment. The best way to visualize this in your case is to imagine the part sitting flat on datum feature A and then a collet-like device--to simulate datum B--contracting around the OD until it just touches the high points. You then measure the hole positions from the center of the datum B simulator, not the center of the part itself.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[kakalee1]

Thanks Dave, Powerhound and others,
It is pretty clear to me now.