Voelcker article written in 2002 3

Herb was making an argument in support of a means to distinguish the three cases. If he picked the first one, "a", there would be little motive to continue and "b-2" is unappealing to the cause.

I think any "b" case is not supportable. I don't recall anything from the 1994 version, which was applicable at the time, suggesting that the RFS simulator would change location and/or orientation relative to earlier references in the DRF. In fact, the opposite is true, with specificity that the centerplane in such an example remain exactly aligned - Fig 4-15, '1994.

There was no need to rotate the relative position of the hole that is being controlled in the diagram. If the diagram is redrawn with that hole directly vertical above the datum B feature, then the movement of the datum C simulator is clearly not expected.

Ultimately it's a case of "Can't sell a solution if you can't show there's a problem." It would have been trivial to state paragraph and figure to support b-1; not sure why he chose not to.

Further clarification is found in Tandler's patent COMPUTER AIDED DESIGN (CAD) SYSTEM FOR AUTOMATICALLY CONSTRUCTING DATUM REFERENCE FRAME (DRF) AND FEATURE CONTROL FRAME (FCF) FOR MACHINE PART
See Section 8, lines 15-24.

I believe Voelcker's argumentation for figure (3b-1) was coming from the wording of para. 4.5.3(d) in Y14.5M-1994 and from fig. 4-2 in Y14.5.1M-1994. The paragraph basically says that the tertiary datum feature simulator (called True Geometric Counterpart, TGC, in that version) must only be oriented to the primary and secondary TGCs. And the figure in the math standard literally shows that.

Unfortunately, figures in the Y14.5M-1994 standard (4-7, 4-9 and 4-15), as well as applicable figures in Y14.43-2003 for gages and fixtures, clearly show the center plane of the tertiary TGC aligned with the axis of the secondary TGC. So in short, before 2009 there was a conflict in the standards.

According to para 4.11.4(e) and indirectly to para. 4.5.2(c) in Y14.5-2009, in addition to basic orientation, the tertiary datum feature simulator must have basic location to the primary and secondary datum feature simulators. With that defined, there is no longer a conflict between the text and the applicable figures in the 2009 version (4-9, 4-15 and 4-18) of Y14.5-2009.

So if the discussed article was written after 2009, the correct answer would be (3a). If, however, someone wanted to override the basic-location requirement, the translation modifier for the tertiary datum feature would have to be used. And then the correct answer would again be (3b-1).

3DDave and pmarc,
Thank you for your clarification.

Pmarc,
Speaking about the gaging standards ASME Y14.43 (2003 and 2011, for Y14.5-1994 and Y14.5-2009 respectively) may I ask you a follow up question:

Why do you think that in Fig B9 from Y14.43-2003 datum feature B simulator requires a minimum two points of contact?
I would expect --- because the 2003 gaging standard ( Y14.43-2003) is written in support of ASME Y14.5-1994 --- that the minimum points of contact for the tertiary B to be one point of contact and not two. Again, we are talking about 1994 where orientation (and not location) is the "official" (text written requirement).

If you see, the equivalent figure in 2011 gaging standard (Y14.43-2011) Fig B-9 page 82, stay unchanged versus 2003 gaging standard, even the datum feature simulators requirement have been dramatically changed from 1994 to 2009 ( orientation only versus orientation and location).
Why do you think that no update was needed for this figure ( B-9) in 2011?
Am I missing something?

Moreover, I have noticed that Fig B-12 page 92 (Y14.43-2011) does not have the same requirement ( two points contact) as Fig B-9 page 82 has (again, I am talking about minimum of two points contact). Is this an intentional omission? Should I assume the same two points minimum requirement applies regardless if specifically shown or not (veriage/ text adjacent to the figure)?

Thank you

pmarc - with customized datum frames, per Voelcker, all three are available, which is exactly what he got into the standard.

aniiben - having Y14.43 is a huge distraction to give gage makers a different basis to work from than engineers expect to happen. It won't get withdrawn, so it adds yet another means of creating conflict for the foreseeable future.

aniiben,

Correct me, if I am wrong, but your statement that the minimum number of points of contact for the tertiary planar datum feature should be one and not two seems to stem from the 3-2-1 rule. While the rule is true in cases where all datum features are planar, it does not always work that way. I am afraid Y14.5M-1994 doesn't address that, but in 2009 version of the standard there is a figure showing similar datum features configuration. It is fig. 4-30(a). In the "Means this" part of the figure it is (somewhat enigmatically) said that the datum feature simulator progresses from 15.1 to 14.9 normal to make the maximum contact with datum feature B. The "maximum contact with datum feature B" means that one has to try to bring the simulator B as close to the datum axis A as possible. For the actual geometry as shown in that figure the maximum contact with datum feature B will be when the simulator and the datum feature contact at two points. With one point of contact further progression of the simulator towards 14.9 would still be possible.     

Also, keep in mind that the requirement for the datum feature simulators to be basically located relative to each other does not apply to cases where secondary or tertiary planar datum feature referenced at RMB has a locational relationship to the higher order datums. In that case the simulator for the planar datum feature is movable (adjustable in size, as they call it in para. 4.5.2(f))



3DDave,

I am not sure what you mean by "customized datum frames, per Voelcker", but I would say that in the Y14.5-2009 realm there is a way to specify all 3 cases from Figure 3 without using a customized DRF concept as defined in paras. 4.22 and 4.23.

Case 3a is simply |A|B|C|.
Case 3b-1 is |A|B|C[translation]|.
Case 3b-2 is |A|C|B[translation]|.

So in that context his complaints that designers don't have tools to define those cases don't apply any more.

pmarc said:

While the rule is true in cases where all datum features are planar, it does not always work that way. I am afraid Y14.5M-1994 doesn't address that

Click to expand...

pmarc,
Should I understand that, in your opinion, within 1994 (since there is not much clarity on this subject) the maximum contact point requirement is non-existent? (lets pretend we are talking about fig B9 configuration from 2003 gaging standard, but drawing defined as per ASME Y14.5-1994)?
In other words, I COULD have an interpretation of one point contact (as specified by aniiben/ OP) being legal or valid? I am thiking this because, since 1994 does not address it, and only 1994 standard is shown, not many users go to the gaging standard (2003), therefore, the one point interpretation could be argued, don't it? I guess, based on your statement above, there is nothing within 1994 to prevent one point interpretation on fig B9. Please confirm or invalidate my comments.

Maybe this is a part (granted...small part) of the "tertiary datum problem" which prompted the revision in 2009, don't you think ? (one point versus minimum of two points of contact)

This is a diagram that shows the b-1 figure argument is carefully contrived to have an extra degree of control that is beyond being a tertiary datum control. b-1 depends on the tertiary datum being suitable, all alone, as a feature sufficient to provide reliable orientation control with minimal sensitivity for the part based on acceptable variations. So here's what happens when the depth of the feature is changed without changing the allowable variation; noting that depth is not mentioned in the tertiary datum argument.

greenimi,
This is just pure speculation, but I would think that many people could claim that a single point of contact according to '94 version of Y14.5 is the correct solution. I would also say the arguments for minimum-two-point-of-contact interpretation could be provided on an extension of principles basis, but, as you pretty well know, for many people that wouldn't be enough.

3DDave,
If a feature by its nature (for example due to its size) isn't good candidate for a datum feature, then it should not be chosen a datum feature in the first place, because it will most likely not function as a robust datum feature in a real design.

And my apologies, but I have to say ask for clarification again... Could you help me understand what the diagram is showing?

pmarc - are you saying that detents for indexing don't work? Because the dents that are used for indexing cannot by themselves fix the rotation of the part.

It's as clear in this diagram as it was in Voeckler's paper and is just as good a candidate. I could drop a shear key into such a shallow notch and provide huge levels of fixity, even without leveraging against the bottom face. If a change in an unmentioned characteristic can cause the feature to become not a good candidate, then it isn't my fault that it was overlooked as a flaw in the original argument. That those making the argument went to the math standard and undermined the main standard was unfortunate.

3DDave,

Not to pile onto pmarc's question but could you also expand a little upon what you mean by the below? It seems to me that in many cases a tertiary datum will be able to have at least one extra degree of control beyond what it is technically "allowed" to constrain as a tertiary datum feature - and/or at least one more than is needed to constrain the part (technically over-constrained) which is the heart of the tertiary datum problem. I think the key is in what you mean by "sufficient to provide reliable orientation control with minimal sensitivity for the part based on acceptable variations" but I'm just a tad unsure what exactly what is meant by that.

3DDave 7 Jan 19 20:21 said:

the b-1 figure argument is carefully contrived to have an extra degree of control that is beyond being a tertiary datum control. b-1 depends on the tertiary datum being suitable, all alone, as a feature sufficient to provide reliable orientation control with minimal sensitivity for the part based on acceptable variations

Click to expand...

3DDave,
What I was actually trying to say was that I didn't really understand the point of the argument you were making. And the diagram you prepared didn't help me to understand it.

For example, it wasn't clear to me why the dimension value on the left picture was greater that the dimension value on the right picture.

But you don't have to explain it to me now. I see the discussion isn't going in good direction, and I'm not interested in another tense debate with you that in the end won't help anyone on this forum.

pmarc - Had I shown my diagram at the time, it's very likely that the committee would have rejected datum mobility as the default interpretation for the Math standard. Instead they were shown a very special case with unstated assumptions. Understanding those assumptions is helpful.

Pmarc,
Yes, my mindset was on 3-2-1 "rule" per the planar datum feature. Thanks for your input.
I also went by the degrees of freedom (constrains) approach. The only degree of freedom left for B (as tertiary) in Fig B9 is a rotation and "this rotation" could be stopped with only one point of contact.

I did not understand why in order to arrest the last rotation two point of contact is are needed.

pmarc said:

The "maximum contact with datum feature B" means that one has to try to bring the simulator B as close to the datum axis A as possible.

Click to expand...

That definition doesn't work so well for Fig. 4-31(a).


pylfrm

pylfrm,
Agreed. This was an ad hoc definition created for the purposes of fig. 4-30(a).

In fig. 4-31(a) "the maximum contact with datum feature B" may exist for example when the simulator B is 5.1 from datum axis A, and pushing it further towards A will only make the contact "less maximum", so to speak.

I guess that is the reason why people complain about this figure. I can imagine that with the use of a physical gage it would be very hard to determine the very moment of maximum contact between the actual datum feature and its simulator.

From 4.11.10 Translation Modifier - what does this mean : "the datum feature simulator is able to translate within the specified geometric tolerance"

What geometric tolerance applies and where is it specified?
What does it mean for a datum feature simulator to be "within" a tolerance?

The "means this" of Fig 4-32 indicates no limitation of any kind on the translation of the datum feature simulator.

---

4-31 (b) is even more of a problem as half the profile tolerance zone lies on the far side of the fixed datum feature simulator. And what purpose is there in making the location of the profile zone leader change and the location of the datum symbol for B change? They could have been the same as (a) and (c).

I'm a little surprised they did not suggest using the candidate datum method from Y14.5.1 as suggested by 4.11.2 for 4-31(a) If the surface is convex then there is no 'maximum.'

In fig. 4-32(b) the center plane of the datum feature simulator B is able to translate within position tolerance applied to the datum slot B, that is within a distance of 4.9-5.1 from datum axis A. Technically, it could translate further up or down, but once it gets closer than 4.9 to or farther than 5.1 from the datum axis A, the distance/separation between two expanding parallel surfaces constituting the simulator B will no longer be maximum possible for given actual geometry of the slot, and that will be a violation of RMB condition.

Per para. 4.11.2, the candidate datum method is always available for special cases of actual datum feature geometries. I agree that in case of a convex surface in 4-31(a) there will be no maximum contact. And I think the same applies to fig. 4-30(a).