Datums and MMC question

[chrisr1980]

Hello,
I am new to this site and I hope I can explain my question clearly. I have a cylinder that has 2 holes on a surface. Datum A is my plane, Datum B is the outer diameter of the cylinder, and datum C is the diameter of one of the holes. I have a true position callout on Datum C to datum B with a tolerance of .007" MMC. The 2nd hole is located to Datum C with a true position tolerance of .014" MMC. All diameters are +/- .002". When using a vision system, the true position of the non-datum hole to A/B/C is reading up to .030". However, when using a functional gage that allows for the MMC on Datum A and B to move freely, the part passes. Even with all of the bonus tolerance you can never get to a .030" reading that would pass, but in looking at figure 7-18 of ASME Y14.5-2009 and reading 7.3.6.2.2(b), I believe that the functional gage is correct in checking the part. The gage is set up to with Datum B machined to the MMB, a pin for datum B located in the basic dimension location +/- .0002" sized at its MMB condition minus the true position tolerance, and the non-datum pin located at its basic dimension location +/- .0002" sized to MMB minus true position allowance. Is the functional gage properly checking the part? Is there anything I can add to better clarify my question?

Thanks,
Chris

 

[ALK35]

Chris,

If you can attach a picture may clear a lot of thngs up for me. There can be no MMC condition on Datum A since it is a plane. Is there a perpendicularity tolerance on Datum B ar MMC to Datum A? Is Datum C relative to Datum A&B or just B and at what material condition? When using the vision system how are you setting up the part? In the functional gage Datum B would not be a pin, Datum C would be a pin. Again, if you can, please attach a drawing.

Thanks

ALK

 

[pmarc]

Yes, a sketch would really help. Though I think I picture the geometry correctly, I got confused when reading how the part was dimensioned.

 

[fsincox]

I believe surfaces can now have MMB per ASME Y14.5M-2009

 

[pmarc]

Frank,
Surface datum feature can now be modified at MMB provided that this surface is anyhow located relative to higher precedence datums, as shown in figs. 4-29 through 4-31 in Y14.5-2009.
The concept will not work if a face is primary datum feauture or is only orientationally related to other datums (like for instance a DRF constructed from a typical set of 3 mutually perpendicular side walls of a block).

I think in this particular case here there is no need to employ 'surface-at-MMB' idea. But again, a sketch with actual dimensioning scheme would really help to answer OP's questions.

 

[chrisr1980]

Attached is a sketch of the part. Our functional gage in this case is 1.002 to contain the OD, with a .141 pin located at the basic dimension for datum C and a .134 pin located at the basic dimensions .200, .250.

 

[pmarc]

Datum feature simulator B for containing the OD shall be dia. 1.004 = 1.002 (MMC size) + 0.002 (perp. tol).
Other pins are correct.

 

[Belanger]

It looks like this drawing uses the pre-1994 system, so don't worry about the MMB stuff on a surface. But a drawing sure helps us understand the question better!
Perhaps the vision system wasn't accounting for the extra "fudging" that comes from having datums B and C referenced with the M modifier (what is called datum shift, or shift tolerance). I didn't bother to do the math. But a functional gage DOES account for datum shift, so I'd go with the functional gage results anyway.
The disadvantage of a functional gage, however, is that it doesn't give you variable data, which is often desired in today's inspection rooms.

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

http://www.gdtseminars.com

 

[powerhound]

Out of curiosity, why are the holes called out and shown like that? Is there are reason they aren't shown in a horizontal or vertical orientation with only the spacing and a single dimension from the center of B thus no datum C necessary?

Powerhound, GDTP S-0731
Engineering Technician
Inventor 2010
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[chrisr1980]

pmarc,
I have checked the tool and it does have the .002 added for the perpendicularity.

Belanger,
I agree with you on this point as well. Our supplier of this part has not been able to hold a proper CpK when taking variable data so we have switched to 100% inspection on these features while a new supplier with a superior machining method is brought on line.

Thank you to everyone who helped me to understand this issue.

 

[pmarc]

chrisr1980,
Of course keep in mind that gages also have their own tolerances (for pin sizes and for geometrical relationships), so in fact real gage used for inspection of the part should differ a bit from the theoretical gage we talked about here.

 

[fsincox]

pmarc,
Good point, Thank you for the clarification.

 

[ALK35]

Chris,
I took your part and showed what would pass on a functional gage, and then took that part and set it up at RFS to Datum B and Datum C. You can see that a part that would fit on the gage would seem way out on a system that does not applied the datum shift.

 

[chrisr1980]

pmarc,
We are well aware of this and believe it will be the next discussion point with our end customer. Our plan is to draw the gage to nominal and overlay a dxf of the dimensions taken from our measurement system. Based on initial testing we have found that parts that would be out of tolerance are fitting in the "gage".

ALK,
Do you have a drawing showing this or data for how far out the true position could be RFS that still passed?

 

[ALK35]

I will try again to post them

 

[ALK35]

I can only do one at a time

 

[ALK35]

Here is the last one

 

[pmarc]

Nice clarification, ALK35. Your observation confirms datum shift's nature - it allows toleranced hole to be more out of true (ideal) position than if B and C were referenced at RFS (RMB according to 2009 wording).

I just really hope that based on this pictorial explanation no-one will ever conclude that 'datum shift' always has negative influence on part's functionality - that is not true. If OP's part function is simply to assemble the cylinder with a counterpart (being a pocket with two pins for example), greater locational error of toleranced hole will not have any impact on the assembly. Parts will still mate since functional boundary of the hole (called virtual condition) will never be violated by hole's surface, even if datum shift is maximum value possible.