Profile and position callout 1

[qcgoofyfoot]

I have a simple rectangular part, 6"x4"x.125" thick with a hole in the center.

The flat surface is called out at "A".

The 6" edge is "B".

The 4" edge is "C".

Profile of the perimeter is called out to be .01 with respect to datum A.

The position of the hole is called out with respect to ABC.

Is this a valid GDT callout?

Specifically, is it acceptable to callout the perimeter with a profile, and then use the edges as datums to find the hole?

The perimeter is not being called with respect to ABC, so I think it is ok.

 

[Belanger]

Yes, it's OK. The profile tolerance controls the form and orientation of the two sides.

Realize that the .01 tolerance is not carried over to the position tolerance. (When the position is checked to ABC, those three datums are formed by high-point contact, even if B and C are wavy or slightly angled.)

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[qcgoofyfoot]

Thanks for your reply John-Paul.

I just wanted to make one thing clear. The profile callout for the perimeter of the part is .01 all around. The profile callout is not from one datum to another. Is this still ok?

I know that it is would not be "legal" to call out the profile .01 ABC all around.

Thanks again!

 

[Belanger]

Yes -- all is A-OK and the answer wouldn't change.

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[axym]

I'm not sure that it would be illegal to call out the profile to ABC all around. This would force A to be flat within .005 and B to be perpendicular to A within .005.

I'm not recommending this practice, but I don't think it violates any Y14.5 rules. What do you think, John-Paul?

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Hmmm -- illegal might be too strong of a word, but I don't think it would add any value.

The corners are already implied to be 90º so the perpendicularity of the four corners would automatically be controlled to .01 by virtue of the "all around." Referencing B and C in there would almost be circular logic since two of the edges are forming B and C. And the flatness of datum feature A is not affected at all.

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[fsincox]

axym,
I too am under the impression that it would cut the iolerance to the datums as you say, anyone else?

 

[Belanger]

Do you guys think that restraining the perimeter to ABC cuts the tolerance in half? I don't think so -- I'll have to play around with a sketch and think...

How about this: By referencing ABC, what additional quality would it control? Size, location, orientation, or form? Those four qualities were already tackled by having profile all around to datum A. And the effective tolerance value wouldn't change because the datums that are created are from high-point contact anyway.

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[dingy2]

The original concept with profile of a surface to datum A looks fine as long as the leader line indicates "ALL AROUND".

If one referenced the profile to secondary datums B and C in addition to the primary, would one check datum B and C or assume they are perfect? One could always say that rule #1 would control the form including datum B and C surfaces but I think we had this discussion before.

One would have to assume datums B & C were perfect thus we are not checking the full perimeter only two (2) other sides relative to A, side datum B & end surface datum C.

I think the original method is best.

Dave D.

http://www.qmsi.ca

 

[Belanger]

Thanks Dave,
I agree -- and this saves me from making a sketch, which I always seem to goof up when it comes to posting one on the forum.

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[fsincox]

I was not intending to recomend that option I was just trying to verify with others, the valididty of the reason I would not recommend it. I do believe it is forced to be correct. If you have a simple square part and one side or two are datums and a profile is used all around the opposite sides can only stay in their required tolerance band if the datum surface only veries inside of the true profile.

 

[SeasonLee]

The original post of profile callout with only one primary datum is absolute correct and I don’t think we need secondary and tertiary datum on the all around profile callout, the reason is to avoid profile specifications that are confusing, Alex Krulikowski mentioned Profile Datum Rule in his book Advanced Concepts of GD&T Chapter 22 : A profile control should not be applied to the surfaces it references as datum features, he says it’s a common sense rule since confusion will exists on how to interpret the profile tolerance zone in the datum contact areas with the features, please look at fig 22-10 for detailed information.

SeasonLee

 

[axym]

I still think that the tolerance for the datum features is effectively cut in half. Dave, we wouldn't ignore datum features B and C or assume that they're perfect. Rule #1 doesn't enter into it either.

When only datum feature A is referenced, we can translate and rotate the zone to try to get everything in. When datum features B and C are referenced as well, the zone is the same but it must be oriented and located to the datums. The datums are on the high points of the datum features, so the datum features can only use half of the zone.

This is hard to convey in words, so here's a diagram showing the two scenarios. Again, I'm not recommending that this technique be used but here's what it would look like if it was used.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

 

[dingy2]

I agree with SeasonLee reflecting Alex Krulikowski's thought on profiles and Evan's diagram referencing profile of a surface all round to A, B and C just doesn't make sense. It is confusing.

Whenever we have a datum surface that does not have a geometric qualifier, we must assume that datum surface is perfect and as per section 4.10.1 (2009 edition) we use a 3 point contact on the primary, 2 points on the secondary and 1 point on the tertiary. This hasn't changed since I was measuring parts so many years ago. Of course, we could look for the high points on the surfaces but that only happens when checking fixtures are used and not on a CMM. Do you think a CMM Operator will scan a surface for the high point and then try to develop 2 other points to stabilize the part? Don't think so.




Dave D.

http://www.qmsi.ca

 

[Belanger]

I still disagree, Evan

The profile tolerance zone is equal bilateral around the "true profile." You are making the datum planes the true profile, and that's not the case. The true profile is a perfect shape -- and here's the key -- at the prescribed distance from the datums as given by basic dimensions.

Are you saying that the basic distance from datums B and C to the true profile are zero? But datums B and C are formed from the high points of the actual edges. Thus we have a circular argument, and I say we should leave off B and C altogether.

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[axym]

Dave,

I agree that referencing the profile all around to A, B, and C is confusing and I wouldn't recommend doing it.

But I'm not sure where you got the idea that we must assume the datum surface is perfect. The datum feature simulator is nearly perfect, and the datum is perfect. Here's what the standard says in 4.10.1:

"Where a surface is specified as a datum feature, the high point(s) establish a datum plane".

That's the principle. When an imperfect datum feature interacts with a perfect simulator, the contact is at the high points. All the other stuff about the number of points of contact is derived from this principle.

This isn't the same thing as picking 3 random points on the primary, 2 on the secondary, and 1 on the tertiary. If the CMM operator does that and therefore doesn't find the high points, then it's not the same datum reference frame! Maybe that's what you mean about assuming that the datum feature is perfect. It would have to be perfect for the random 3-2-1 method to get the proper datum reference frame. I would agree that most CMM operators would use the method that you described and not find the high points. But that doesn't make it any less wrong. I was a CMM programmer once myself, and learned the lesson of high point datums the hard way.

John-Paul,

Am I making the datum planes the true profile? Yes, I think that makes sense.

Am I saying that the basic distances from datums B and C to the true profile are zero? Yes, I'm fine with that as well. Datums B and C are formed from the high points of the datum features. I don't think it's a circular argument.

I've attached a closeup of the diagram, with the different pieces of geometry labeled. The true profile (purple solid line) and the datum (dashed black line) are coincident. Do you think this makes sense?

It's too bad that Y14.5 relies so heavily on the special case of three mutually perpendicular planar datum features as an example. It masks some of the subtleties of datum reference frames because the true profiles, the datum planes, and the planes of the datum reference frame are all coincident.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

 

[drawoh]

I agree that referencing the profile all around to A, B, and C is confusing and I wouldn't recommend doing it.

Using A and B as profile datums makes sense to me. Datum_C, the tertiary datum, is defined by the perimeter, so it cannot control the perimeter. B also is defined by the perimeter, but it provides control over the 4"_sides.

JHG

 

[dingy2]

Evan:

You said ""Where a surface is specified as a datum feature, the high point(s) establish a datum plane".

Since you were a CMM programmer, please reveal to us how one would set up with a high point using a CMM without using a 3 point set up on a plane as described in 4.10.1? There is only 1 high point so take it from there. How would one set it up in real life?

Dave D.

http://www.qmsi.ca

 

[Belanger]

Evan,
On your last sketch, how is it that the true profile just happens to also be the same as the high points? If the surface hovered near the upper limit of the zone, then the datum plane would move up there. Would your true profile also move?

Also, would any of this discussion change if we were looking at a single flat surface (rather than four sides) and use profile of a surface to itself as a datum? I think most people would see the problem with that. So I'm trying to figure out what's different when there are four sides...

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[axym]

I sense that we're touching on something fundamental here.

The true profile didn't just happen to be the same as the high points. If the surface hovered near the upper limit of the zone, the datum plane would move, and the true profile would also move. The datum plane follows the high points, and the true profile follows the datum, and the zone follows the true profile.

If we were looking at a single flat surface with a profile to itself, it would seem even more ridiculous and more people would have a problem with it. But I don't think that the geometry of the datum feature, simulator, and datum would change at all.

To me, it's another technique that is technically not illegal but could be considered a bad practice. It's kind of like saying "the allowable gap between the surface of my foot and the perfectly flat floor I'm standing on must be zero plus or minus 0.25 mm". :^)

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca