Profile to replace Position/Size

[[Ja...]]

I have a large radius (>3") on a part but I only have about 5degrees of arc. We are supposed to hold +/-.005 on size and a .005 Position. We cannot measure this on anything we have.

I am suggesting a composite Profile control of .005 to the same DRF and .010 to no DRF. Will this accomplish the same thing?

[[Br...]]

Please sign in to view this quote.

Position tolerances are ONLY for features of size(assuming you're working to Y14.5). Profile is a good way locate surfaces such as small arcs.

Regarding the composite Profile tolerances, the lower segments are refinements of the segments that precede them, and therefore must have a smaller tolerance. Think about it, if you already control the profile to a DRF within .005, then it has to by necessity at least, be within .005 with no DRF. If it passes the .005 profile tolerance aka "PLTZF", it will always pass the .010 lower segment, aka "FRTZF".

[[Ja...]]

Yes Brett, I am working to Y14.5.

The refinement information makes sense, but if I held a Profile of .005 to guarantee Position and size, would I not be taking away allowable tolerance on my size? Can I use .005 Profile to datums as a refinement on Profile of .010 with no datums?

[[Br...]]

No, whatever datum feature references are found in the lower segments, must repeat the the upper segment in the same order of precedence. It doesn't work the other way around. That wouldn't be a refinement, you would be adding constraints instead of taking them away.

If you're tying to roll whatever would've been a feature of size with a size tolerance and position tolerance into a Profile of surface, first you would want to look at where the surface boundaries would have existed. On an internal feature, Your inner boundary for your profile would be equal to the virtual condition of your original tolerances, and the outer boundary would be equal to the Resultant condition. Now you have an inner and outer boundary. The distance between the two boundaries would be your tolerance value for the PLTZF. This is really the only way to roll up the size tolerance and the position tolerance into a single value. Then you can make your FRTZF .010 to no DRF. The .010 will be smaller.

[[Ja...]]

Brett,

You are a genius. After looking through Y14.5 and learning some more, that makes a whole lot of sense. Thank you for that.

Now here is the next step, which may be more of an ethics question. The customer wants a hard size and position dimension. If I turn this into a composite profile, internally, can I do the following:

(1) "Scale" my tolerances, such that my original .005 Positional tolerance zone becomes (after PLTZF determination) .020 Profile tolerance zone. The zone is 4x as large. Can I essentially say a .010 Profile result would correspond to a .010/4 = .0025 Position result i.e. Position = Profile/R where R is the ratio between the two zone sizes?
(2) Similar for the size dimension. Map my .010 Profile deviation to a corresponding size dimension, whether above or below the zone.

[[Br...]]

Well I think you really need to tell your costumer that you cannot make it to this drawing. I thought this was an in-house thing where you were working with your own designers.

If the position tolerance is at RFS, then you need to be able to produce the Axis of the cylindrical feature. The Axis is determined by the Actual Mating Envelope. If there are no opposed elements such as with a partial arc, no actual mating envelope can be made, therefore no axis, and no position tolerance. When It comes down to it, the tolerance just doesn't make any sense for this application, and guessing what the designer had in mind here may not be the best approach.

[[Aa...]]

Please sign in to view this quote.

Haha, Brett! A lot of your analysis is spot on. But it seems that paragraph is stuck in the world of Axis Interpretation. From the point of Surface Interpretation, the location of the axis isn't what's critical. But in cases of RFS, the size of the UAME is still needed to determine the size of the inviolable theoretical boundary, so your conclusion about not being able to work from that specification is still supportable.

Besides, the only cmm evaluation option I know of that will evaluate according to Surface Interpretation refuses to do so whenever the angular extent of measurement is too small.

[[Br...]]

Please sign in to view this quote.

at RFS you only have the option of axial interpretation. That's why I specifically made it clear I was talking about RFS. Only at MMC do you have the options to chose between Axial and Surface interpretation. And the surface interpretation doesn't necessarily require an AME. Its just a boundary at MMC that no element of the surface can violate. That's why I made the point about RFS (assuming that is his case.) requiring an axis because I think you could make the case to do some sort of profile tolerance to create that Boundary at MMC. But even then, with a partial arc, the actual surface could move infinity in one direction and never violate the boundary. Its just not the right tolerance.

ASME Y14.5-2009 7.3.2 RFS as Related to Positional Tolerancing

Please sign in to view this quote.

[[Aa...]]

Please sign in to view this quote.

Hmm... the standard's not really all that clear or consistent on this point. Note 1 under 2.8 :

Please sign in to view this quote.

7.3.3.1(a) (which is in the middle of a section about MMC tolerances):

Please sign in to view this quote.

And interestingly enough, in support of your argument, every other reference to surface interpretation in the standard is in context of sections about MMC or LMC.

[[Br...]]

I guess I don't understand where you are saying there is an inconsistency. If you are hung up on the wording in 2.8, those are two different clauses. The first sentence just makes it clear that it is referring to an axis interpretation in the following paragraphs. The next sentence is just saying surface interpretation takes precedence. It's not saying RFS can have a surface interpretation. Its just a completely different thought. So in other words, if you get two different results at MMC between a surface interpretation verses an axial interpretation- The result from the surface interpretation is the correct one.

The reason you only see surface interpretation under MMC and LMC is because it's the only way that it makes sense. With MMC for instance you can effectively take the tolerance off of the axis, and place it on the physical surfaces becouse you can calculate a virtual condition boundary. The point is to simulate hard gauging. With RFS you cannot calculate a virtual condition boundary becouse it is regardless of feature size . There's no way that you could make the case for a surface interpretation at RFS. It just doesn't make sense to do it that way.

[[Aa...]]

With Axis Interpretation, you're at the mercy of the instability in the orientation of the UAME. (For example, in situations of holes that are tapered in reality, the axis of the UAME at the more open end can end up considerably skewed from the derived median axis, especially where the form control is not significantly tighter than the position control.) Surface Interpretation overcomes this issue. This is true for MMC, LMC, and RFS.

We regularly use commercially available software that will toggle between axis and surface interpretation for any position requirement, and we have a healthy appreciation for the benefit of surface interpretation, even for RFS.

[[Br...]]

Please sign in to view this quote.

So how do you do a surface interpretation with RFS then? If you are designing the software, how is it calculating a surface interpretation at RFS? If you create a virtual condition boundary, you've just changed your tolerance to MMC. Surface interpretation just completely goes against the definition of RFS. If the tolerance is REGARDLESS OF FEATURE SIZE then the size of the feature can have no bearing on the location of the feature. With surface interpretation, the larger the size(of an internal feature at MMC), the more the part can shift around without violating the VC boundary. Maybe if you had some sort of dynamic boundary that expands with the size but that's a stretch that will introduce other problems, plus there's just no support for something like that in the standard. You'd be making up your own rules.

[[Aa...]]

The boundary is at basic location and at the UAME size +/- the position tolerance (depending on internal or external.) That's why I said the part at hand is still impossible to measure as drawn. Without enough angular extent UAME size is a joke.

[[Ja...]]

Please sign in to view this quote.

If I'm not mistaken, you two are describing the same boundary.

[[Br...]]

Please sign in to view this quote.

Were does the standard describe that? Y14.5 defines it as having a boundary existing at the virtual condition. RFS has no virtual condition. What you describe is just sort of made up. There is no support for that interpretation. I think you already pointed out the flaws in that type of practice. It's probably why we don't see something like this already introduced in the standard.

[[Aa...]]

We're not describing the same boundary. Brett is talking about offsetting the MMC size by the tolerance, and I'm talking about offsetting the UAME size by the tolerance.

Actually, the Y14.5.1M-1994 goes into Surface Interpretation on RFS in great detail, including a table outlining how to calculate the boundary size for each material condition.

table.JPG

[[Aa...]]

Please sign in to view this quote.

It's not made up. It's in Y14.5.1M-1994. (That's the mathematical definitions standard.)

[[Ja...]]

Brett,

Isn't a surface interpretation at RFS exactly how you described to set up a Profile callout that would encompass both Position and size? The location of the feature, let's assume internal cylinder, is not awarded more Positional tolerance as the size changes. But the size can change through direct tolerance and will have an effect on the outer/inner allowable boundary that must contain the surface as the feature moves throughout it's allowable Positional tolerance zone. This would be a surface interpretation at RFS, no?

Y14.5-2009 Fig. 2-14 shows Inner and Outer Boundaries @RFS for an internal feature.

[[Br...]]

Please sign in to view this quote.

Yes it is. and I shouldn't have used the terms virtual and resultant condition because its really not that. They are just inner and outer boundaries. It's sort of the best you can do to roll that tolerance up into a profile tolerance though. The way it stands, it cannot be verified because it violates the standard. But that's why I said you need to talk to your costumer before doing something like that because you don't really know what the intent there is. Those tolerances should represent different limits and fits, and we don't actually know the function here.

[[Br...]]

Please sign in to view this quote.

Now that is interesting. I don't have an answer for that. All I know is what Y14.5 (the governing document for a particular drawing) has to say. If you have some GD&T books written by various Y14.5 commiteemen, see what they have to say if you don't want to take my word for it.

Alex Krulikowski in his book "Fundamentals of Geometric Dimensioning and Tolerancing says:
"Where a position tolerance applies at RFS, it is interpreted as an axis/center plane tolerance zone. Where a position tolerance applies at MMC or LMC it is typically interpreted as a boundary tolerance zone. "

Don Day in his book the GD&T hierarchy says:
"When a position tolerance applies RFS, the tolerance is controlling the center (center point, axis or center plane). However, when MMC or LMC is applie to a position tolerance there are two different ways to "interpret" the meaning..."

[[Aa...]]

Isn't that the nice thing about committees (especially fluid ones): what one member has to say is never the same as what the whole committee has to say.

[[Aa...]]

Please sign in to view this quote.

[[Ro...]]

Sorry i'm late to the party, so ill be short as my answer may be ret###ed.
So i've come across this problem years ago at a different employer, and since the part was still in "pre production/ first run" status we had our engineer contact the customer engineer and changed the print on several areas that had radii of different sizes blending into each other changed into a profile with slightly tighter tolerance than the XY&Z method would allow.
That was our concession to get the deal done, and since the profile area was relatively easy for us to hold, everyone was happy in the end.
So my answer would be Yes, but you would need customer approval.

[[Br...]]

Please sign in to view this quote.

I don't think this is a controversial topic among the Y14.5 committee. I think you'll be hard pressed to find any of these guys disagree with what the standard already says--that RFS requires the axis or centerpoint to be within its tolerance zone.

[[Aa...]]

Note 2 in 2.8 says that even though the principles are described in terms of axis interpretation, where that's not exactly equivalent to surface interpretation, surface interpretation takes precedence.

Surface interpretation is not fully explained anywhare in Y14.5. That would be redundant, since Y14.5.1 fully explains it for RFS, MMC, and LMC, and 1.1 of Y14.5 invokes Y14.5.1.