Profile with no datums 1

[PRuggiero]

Guys,

I've read a couple posts here about surface profiles with no datums but I'm still a little confused. I found this post by Jim (MECHNORTH):

By "best fitting" one can shift the workpiece about its 6dof to find the best fit of the workpiece within the maximum & minimum boundaries (i.e. offset geometries from the basic dimensions) to qualify the part.

So if I had a simple block part, datum A is a face, Datum B the longer edge and C the short edge and I had a profile callout on say the edge opposite of datum B without referencing any datums how would it get inspected?

From the post above it seems that if I placed the surface representing datum B down on a granite table and measured the distance that the basic dimension gives to the surface I am profiling, this would give me the "line" that my profile tolerance would be centered on, correct? Then I am allowed to move the part however I want to get that surface to fall within the profile tolerance.


Thanks,
Pete

 

[pmarc]

For your example this would mean you could simply probe the toleranced surface to see if the difference in reading was not greater than profile tolerance value. But the block could be set up in any way you'd like - on a longer surface, on a bottom surface, etc.

Lack of datums in profile FCF means no instructions on how the part has to be set up for profile measurement. In your case, if it was profile of a surface callout, you would simply measure flatness of the surface.

 

[PRuggiero]

pmarc,

that was my original understanding but what MECHNORTH seems to be saying is that isn't the case. The profile still controls the size because the profile tolerance is centered on the basic dimension in the drawing. This is what is confusing me as he seems to be saying, you need to find where the surface is supposed to be (by the basic dimensions) and then you can use the 6 DOFs to make the surface fit within the profile.

 

[pmarc]

Nope, I guess this was not Jim's point, otherwise he wouldn't say that all 6 DOF's were still available.

If there is no datum in profile FCF, the basic dimension between datum feature B and toleranced surface does not apply = the distance (size) is not controlled.

 

[ptruitt]

Occasionally, parts are freeform or have no mating parts. Perhaps one single profile callout referencing a CAD model might be enough information to control the part. But maybe that part is molded and one gate will be trimmed off. An added profile tolerance might allow the gate to be hand-trimmed. Some here might argue that the molded form could be Datum A and the hand trimming could reference Datum A. But an interpretation (proper interpretation?) of "simultaneous requirements" (ASME Y-14.5-2009 Paragraph 4.19) would imply that both the molded and trimmed areas must simultaneously be within their tolerance zones at the same time and no datums would be needed. The utility of adding datum A, in some situations, may be unwarranted.

Peter Truitt
Minnesota

 

[dingy2]

In my opinion, profile of a surface with no datums is great when the are surfaces that share the same plane but on a 3 dimensional product, the people in Quality do require a datum structure of some sort to measure the part. This would take us back 40 years where there were only a linear/angular tolerances and the personnel in Quality would have to investigate how the part fits and functions to come up with a datum structure.

Dave D.

http://www.qmsi.ca

 

[fsincox]

In the threads on "profile of a line with datums" and then the one on the box with a feature refererenced to a framework being derived by a portion of the feature itself. One of my thoughts was: "How about just getting the part oriented for measurement" as it must be done by the inspector anyway, kind of a helpful start.
Frank

 

[PRuggiero]

pmarc,

Not sure if that is the case, look what MECHNORTH says here:

Getting to the use without DRF is an exercise. 6.5.1 says that it is used to control form or combinations of size, form, location and orientation. Form is not related to datums, therefore in the absence of a drf, all you get is form ... the next extension then is to recognize that "form" for an enclosed boundary is essentially a size control as well. 6.5.4 Indicates that in most cases, Profile requires a drf. (i.e. not always required).

He is saying it controls the size, hopefully we can hear it straight from the man himself.

 

[pmarc]

Profile of a line/surface with no datums works well as a refinement of other profile of a line/surface callout with datums.

Imagine OP's example of a simple block for a moment. Upper segment of a composite profile of a surface callout with datums A, B, C would control location and orientation of the surface relative to referenced datums, and the lower segment with no datums would simply refine form of a surface - in this case, surface flatness.

Some could ask, why not to apply single segment profile of a surface relative to A, B, C and flatness callout separately? I would answer there are sometimes many ways to express the same intent in GD&T we know.

 

[pmarc]

PRuggiero,

I agree, it would be the best to wait for Jim's clarification on that.

My vote is that profile without datums does not control size as long as it is not applied to a surface of a circular/cylindrical feature and the diameter is speciifed as a basic dimension. In this case the size (diameter) is obviously controlled.

 

[fsincox]

Of course it controls the size, indirectly, of opposing dimensions.
Frank

 

[ptruitt]

ASME Y14.5 Paragraph 8.2: "Profile tolerances are used to define a tolerance zone to form combinations of size,form, orientation, and location of feature(s)..." Also: "Depending upon the design requirements, profile tolerance zones may or may not be related to datums." In my opinion, "design requirements" are not "inspection requirements". But it is important to acknowledge that most designers will not have the resources to expect that their parts will be inspected by leading edge metrology labs where entirely free-form parts are being used to define implantable devices, scanned and evaluated against non-dimensioned CAD data using leading edge analysis software.

Peter Truitt
Minnesota

 

[MechNorth]

Consider a 3-D object (any size, shape, etc.) that you can define on a drawing using basic dimensions. If there is no precedence over which feature is a datum feature, what do you do? You use a general profile tolerance or a series of individual profile tolerances, without datum references (you would arguably have simultaneous requirement because of the absence of datum references). What then is your tolerance zone? The profile zone is equal-bilaterally offset from the nominal (basic) geometry, providing a 3-D Max/Min boundary in which the entire part must exist simultaneously. If you're using open-setup metrology, this may mean a lot of extra math for the inspector to figure things out for each feature. If you're using any form of data acquisition metrology equipment (laser or contact digitizers, photogrammetry, etc.) that can establish a point cloud, then you can electronically (manually or using algorithms) shift the inspection data around within the 3-D tolerance zone (envelope) to verify acceptance. This is common practice where CMMs (contact or non-contact) are used; it doesn't mean that everyone does it, or can do it though, just as putting datums on a drawing doesn't mean that the CMM operator uses appropriate simulators to establish the datums (vs minimum-points). In some cases, you can even use scans optical comparators with tolerance-zone transparencies for each view to qualify the part.

So, what I was referring to was a net body, rather than individual feature(s) controlled by a profile control without a DRF. Indeed, a single feature with a profile sans DRF would just be controlled for form (flatness, straightness, cylindricity, circularity, or mathematically defined).

That's not to say that it's ideal in most cases; I have seen it properly used a couple of times, though it could as easily/effectively have been done with datums referenced.

On Pete's OP then, if he had a simple block part, datum A is a face, Datum B the longer edge and C the short edge and he had a profile callout on say the edge opposite of datum B without referencing any datums, then the location of the surface is not controlled, just the form. The product specification would then be incomplete. If it had a +/- size tolerance to locate the face wrt a visibly opposed face, then the profile control would again only be controlling form, but at least the face would have been located.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[dingy2]

Jim, I really hope Designers don't start using Profile of a Surface (all over) without a referenced datum structure. In theory, it might sound doable but in practice, we still need a datum structure or we will create one with the CMM. We do need the 6 dof.

Dave D.

http://www.qmsi.ca

 

[MechNorth]

I definitely don't advocate its use in this way, Dave, but it is legal. Because it is legal, I like to make sure that people understand it fully. Most of us on this forum know that Profile controls are the least understood, and yet most robust control in Y14.5.
Depending on the systems (hardware & software), datums aren't useful or practical in the inspection process. Basic function contact-CMMs typically are used with datums in reasonably-skilled environments, but I've seen a lot of situations where the operators didn't use them at all; it took a lot more data collection, and I'm skeptical of their results, but they were doing it. There's a lot of different data acquisition systems out there beyond the basic CMM. We both have to be careful when suggesting things to people that we consider the entirety of the metrology industry, and not just one specific technology.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[PRuggiero]

mechnorth, thanks for the response looks like pmarc was correct.

Just so I can fully understand; if I had that same block I referred to earlier, but instead had a surface profile of the edges all around with no reference to datums. And it also had two basic dimensions giving the length and width of the part, then that would control form and size correct?

 

[MechNorth]

yes

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[fsincox]

I general, I have always thought of Profile kind of like the engineer's envelope principle on steroids, you get both perfect form at MMC & LMC.
Frank

 

[MechNorth]

yup

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[fsincox]

I actually had proposed at one time as a solution for bearing bore shoulders, position of 0.0 at MMC and 0.0 at LMC to the actual bearing bore. Then someone said: "why not just use profile" (duh, moment)? My management had outlawed profile by then so I just gave up.
Frank