Can Profile -Of -A-Surface have MMC on it's Datum? 4

[CheckerRon]

I can't find anything in Y14.5M-'94 to justify this.

I have been given a part which has an irregular 8 sided shape (not an octagon). The sides of this perimeter are controlled by profile-of-a-surface (individually since their not all of the same tolerance).

There is a prominent center hole (FOS) in the part that is Datum A (primary).

The FCF's of the profile tolerances give the profile tolerance at RFS (as it should) but show the primary
datum A modified MMC.

Is this proper? Can we apply bonus tolerance from Datum A as it deviates from virtual condition to the profile tolerance?

Not sending a sketch. Difficult to do.

 

[KENAT]

Don't send a sketch Ron, they seem to more generate more heat than a well worded question, if only I could learn to word them well.

KENAT, probably the least qualified checker you'll ever meet...

 

[CheckerRon]

OK, One of my colleagues found some words in Lowell Foster's "Geometrics III" where he says that one could properly assign MMC to the datum of a Profile-Of-A-Surface tolerance if the situation warrented it.

Foster cites a cam shaft situation, where the shaft is the datum and the cam profile datum is MMC.

Nothing in Y14.5, but "the guru" speaks?

 

[weavedreamer]

Do you mean like Fig. 4-26 on page 71?

 

[CheckerRon]

I don't know who gave you the star, weavedreamer, but you deserve it. I past right by that figure, looking in other sections.
That is exactly what my part case is.

Thanks,
Ron

 

[weavedreamer]

That eloquent description of yours reminded me of it. Thanks.

 

[ewh]

I think CheckerRon also deserves a star for his excellent description of the issue. As KENAT points out, attaching examples can open a can of worms more often than not.

When the people fear their government, there is tyranny; when the government fears the people, there is liberty. - [small]Thomas Jefferson [/small]

 

[dingy2]

Sometimes the ASME standard is not quite correct and this is one of those situations. There are examples in the standard where profile of a surface is referenced at MMC and is confirmed simultaneously with positional tolerances at MMC.

Profile of a surface should be confirmed in a RFS since it defaults to a range of a bi-lateral tolerance. It can be confirmed by a gauge only if the tolerances were extremely large such as 3 mm but never if the tolerance was 0.4 mm, as an example.

I will absolutely guarantee that if you asked the people in your CMM room how they would approach the datum at MMC on profile, they would state they set it up in RFS. It cannot be measured in any other way.

I have designed a web page outlining this topic. Review it at

http://www.qmsi.ca/datumsMMC.html

Hope this helps.

Dave D.

http://www.qmsi.ca

 

[CheckerRon]

Thanks, Dave.
Good post and good reference.

The part I was reviewing is a like a sheet metal flashing around an off center nozzle on a cylindrical shell. Probably die formed and drawn.

It is only 0.38mm (.015") thick and doesn't have a flat spot on it except for the outer edges that have the profile tolerances on them. The profile tolerances range from 0.2 to 0.5mm, so it falls into the case you describe.

This is what bothered me about the part as soon as I saw it. Trying to envision how one would measure and take advantage of the bonus tolerance offered, as well as whether or not it violated ASME Y14.5.

 

[axym]

Dave,

I wasn't aware that there were any examples in Y14.5 where surface profile is referenced at MMC. Can you point me to the ones that you found?

There should never be any explicit or implied material condition modifiers on the considered feature in a surface profile callout. This is because surface profile can be(and usually is) applied to features that don't have size.

I would agree that it is difficult to reference datum features at MMC using a CMM, but I wouldn't go as far as to say that it has to be set up RFS. There are some software packages that can do it. Another possibility is to use a holding fixture for the CMM inspection. The fixture has a virtual-condition pin for the datum hole that is referenced at MMC. The part can be lined up to the fixture in different ways due to the clearance on the datum hole, but for each one it gets clamped down for stability. The part isn't moving during inspection. The inspector creates a coordinate system in the fixture, and can then try clamping the part in different ways to try get the considered feature to pass.

One last thought. The fact that the characteristic is surface profile doesn't make working with datum features at MMC any more difficult. It would be just as difficult if the considered feature was a hole with a position tolerance at RFS.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[dingy2]

Evan:

Go to page 71 in standard and review fig. 4-26. This is an example of profile of a surface to a datum at MMC. In my humble opinion, it is not an appropriate application and hope the new standard deletes it.

I did say that having a datum at MMC, the CMM Operator will set it up in RFS. I know that one can place a pin of either VC or MMC size but when one is measuring a surface location, we need the exact location of the surface from a base that doesn't float.

Sometimes, one can make a fixture reflecting the surface at it largest allowable size and then have a pin of its tolerance range. Part must fit into its largest size and the pin must not fit between the fixture and the part. This is a bit of a crude measuring method and is only appropriate for large tolerances.


Dave D.

http://www.qmsi.ca

 

[powerhound]

Dave,
I understand your reasoning when trying to check the part in 4-26 using a CMM but I don't believe that means the figure can't be checked and is thus wrong. This is an example of checking the part using a shadow graph and a reticle with the appropriate profile boundaries etched on it. You can rotate the reticle however you want to get the boundary of the part within the boundaries defined on the reticle as long as the 2 hole pattern doesn't violate the MMC boundary. This is actually really easy to do.
I don't see a problem with this dimensioning method as long as the QC department checking it has a shadow graph. Otherwise we are all in agreement that it would be difficult to do without one.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[ringman]

I do not know if it has been stated anywhere within the Standard, but it is my opinion that RFS works well with limited production and CMMs. Whereas MMC works well (Better) for larger production quanties and the functional gages.

This is not to say that either is exclusive.

Opinions?

 

[axym]

Ringman,

I'm a "tolerance based on function" kind of guy. So I would hesitate to connect how well RFS or MMC datum references "work" to different production quantities. But I would make the following statements:

Typical CMM software handles static (RFS) datum reference frames very well, and mobile (MMC or LMC) datum reference frames not well at all. Specialized software or post processing is required to take advantage of the DRF shift properly.

A RFS datum references generally make a physical fixture more expensive, because of the need for adjustable components (collets, mandrels, etc.). MMC datum references are handled easily and relatively cheaply using a physical fixture (fixed size pins, bores, etc.)

If MMC datum references are inspected at RFS, the result is almost always conservative. Borderline good parts will be rejected, because the available DRF shift is not being taken advantage of.

Production quantity affects the economics of inspection and gaging. A larger production quantity may justify making a functional gage that can handle DRF shift, instead of using a simple CMM program that inspects at RFS and rejects some good parts.

Part cost (and the associated cost of scrap) can also affect inspection and gaging strategy. If you're making a small number of very expensive parts and you can't afford to scrap any, it may be worth it to use high-density point cloud scanning techniques and analysis software to take advantage of every shred of tolerance available.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca