Datum Modifier Effects - Planar Surface 1

[Viper555]

Hello,

I have been studying for the GDT Senior level exam (2009) and I'm really struggling to wrap my head around a concept laid out and exemplified in Fig 4-31 through Fig 4-33 (equivalent is Fig 7-34 through Fig 7-37 in 2018 (Reaffirmed 2024)).

The first example, Fig 4-31/Fig 7-34 makes sense because RMB is supposed to represent the feature datum simulator expanding until contact is made to the datum feature, thus constraining the part in rotation.

The second example, Fig 4-32/Fig 7-35, sort of makes sense because the BSC feature requires that it be in contact (similar to that of RMB), thus constraining the part in rotation.

The third example, Fig 4-33/Fig 7-36 starts to really lose me because it's defining the datum feature simulator at MMB, yet it states in Fig 4-33 that is must be in contact with one point. It seems that Fig 7-36 refines this illustration a little along with Fig 7-35, but it still doesn't address my misunderstanding and in fact further confuses me by placing part of the datum feature outside of the LMB/MMB established by the profile tolerance.

Here are my main questions:

1. In Fig 4-32/Fig 7-35, is my understanding of the BSC usage correct in that a datum at BSC requires contact, by definition, to the datum feature?

2. In Fig 4-33/Fig 7-36, why couldn't the part rotate CCW, provided the datum feature falls within the profile tolerance LMB and MMB?

3. In Fig 7-37 (there is no equivalent for LMB in the 2008 standard), why is the part profile able to pass when outside of the LMB and MMB limits?

Any help with clarifying this would be greatly appreciated!

 

[Garland23]

Hi.... First of all, I think you're referring to just Fig. 4-31 and its sub-figures of 4-31(a), 4-31(b), and 4-31(c). They're all on page 73 of that 2009 edition.
Regarding your main question, in that third example where they invoke MMB, yes one point of contact (minimum) is required. The key is that the simulator is built to match the MMB of the surface, which includes the profile tolerance. In all three examples, the part is first contacted with the simulator for datum A (a collapsing diameter) and then rotated clockwise until it makes that one point of contact on B.
I'm not sure what you're asking in question 3, though.

 

[Viper555]

Thank you

Hi.... First of all, I think you're referring to just Fig. 4-31 and its sub-figures of 4-31(a), 4-31(b), and 4-31(c). They're all on page 73 of that 2009 edition.
Regarding your main question, in that third example where they invoke MMB, yes one point of contact (minimum) is required. The key is that the simulator is built to match the MMB of the surface, which includes the profile tolerance. In all three examples, the part is first contacted with the simulator for datum A (a collapsing diameter) and then rotated clockwise until it makes that one point of contact on B.
I'm not sure what you're asking in question 3, though.

Thank you for the assistance and yes I meant Fig 4-31 (a), (b), and (c). Not separate figures like in the 2018 revision.

I'll try to clarify my question 3. Suppose that Datum Feature B comes in with perfect form and orientation at 4.9 in both illustration (b) and (c). What requires the part to make contact with the feature simulator in both of these cases once established on Datum A?

Aren't you left with a .1 gap in (b) and a .2 gap in (c) through which the part can (but not necessarily has to) rotate in order to meet all requirements simultaneously?

It seems to me that illustration (b) and (c) would still permit a datum shift equal to the difference of produced Datum Feature B and Datum Simulator B.

The previous Fig 4-30 even shows that there is rotation permitted at MMB. It would be like in that figure, the part could be produced perfectly at 14.9 and it would have a gap of .2 through which the part could rotate as indicated in the caption. I don't see the difference.

 

[Burunduk]

Considering that Garland's assumption about the figure numbers is correct, I would add the following:

4.31 (b) and (c) work essentially the same way, except the fixed location of datum feature B simulator from datum axis A is at 5 mm (basic) for case (b) whereas it is 5.1 mm for case (c).
In both cases contact between datum feature simulator B and datum feature B is required on at least one point. It is very likely to be just one point of contact in case (c) (MMB).

4.31 (a) is different because the datum feature simulator for B is not stationary. It is allowed some movement to make "maximum contact" with datum feature B. The contact is likely to be at 2 points or even more. Consider that the datum feature simulator is moved from the basic 5 mm location towards datum feature B and doesn't stop at the intial contact, but rotates the part a bit to make more than one point of conract (it is OK as long as it doesn't pass the LMB boundary located at 4.9 mm).

 

[Viper555]

"...In both cases contact between datum feature simulator B and datum feature B is required on at least one point..."

This is at the heart of what I don't understand.

In reference to Fig 4-29(b), the last sentence of 4.16.2 states "Datum Feature B may rotate within the confines created by its departure from MMB and might not remain in contact with the datum feature simulator."

The same verbiage is stated as the last sentence of Section 4.16.4 when referencing Fig 4-30(b).

What is the distinction between the ability to rotate or shift the part in either Fig 4-29(b) or Fig 4-30(b) and Fig 4-31(c)? All of them are using the MMB condition modifier.

Even the equivalent figure in the 2018 revision, Fig 7-36, shows it as not having to make contact.

 

[Dean Watts]

For 2018 Figure 7-37, when evaluating the profile on datum feature B, the part would not be oriented with respect to the LMB & MMB as it shown in the "Means this", it would be rotated CCW, as needed to satisfy the profile tolerance (which references only A), then the separate & independent process of establishment of the datum reference A, B would be as shown in the "Means this". Since the purpose of LMB is to ensure material is present, such as for a subsequent machining of datum feature B, all seems to be OK, I believe.

2018 Figure 7-36 though, I think is an odd one. Since MMB does not make sense for a planar primary datum feature, or for any primary datum feature that is not sufficiently closed to be constrained by the simulator/TGC, then why does Figure 7-36 make sense? I think it does not, so unless somebody points out something I may be missing, I will be working towards the next revision of Y14.5 having an explicitly stated restriction on the application of MMB to features that are sufficiently closed to be constrained by the TGC or location constrained to the higher precedent datum reference frame and also overlapping the origin of the higher precedent DRF (so not an offset datum feature such as datum feature B in these figures). In other words, when MMB is applied, it should only be for cases in which the potential shift of the part relative to the the DRF must be physically limited by either the datum feature being sufficiently closed or, for lower precedence datum features, by a physical constraint between the datum feature and its TGC.

As Burunduk said, the difference between BSC and MMB in Y14.5-2009 was just a matter of where the fixed TGC/simulator is located. In 2018 the intent was to allow datum referance frame shift for situations like the one shown in Fig 7-36, but the MMB case does not make any functional sense, I believe. The application of LMB to a case for which the TGC "does not fully constrain or limit the datum feature" still makes some sense, since it is still ensuring material is present. That said, I think we would be better off to just say that the LMB must remain inside or on (at a low point) the material. I see no reason to require that an extremity of the datum feature remain between MMB and LMB (as stated in 2018, 7.11.11 and 7.16.7).

P.S. - Burunduk - I sent you a message within the Eng-Tips site yesterday. Letting you know about it this way, since I don't know whether or not you will receive a notification.

 

[Viper555]

Wow! Thanks! I think it makes sense now. The conflicting illustrations between the revisions (and the bit talking about the extremeties was throwing me off).

Just to make sure...you are saying that no matter what the scenario is, the entirety of Datum Feature B still must fall within the profile tolerance of 0.2 and you ARE able to rotate when a modifier is present whether it be BSC, LMB, or MMB since the rotation is only controlled by Datum Axis A in this case (RMB would lock it in).

CW rotation locks it at the high points (MMB and BSC)
CCW locks it at the low points (LMB)

If you make the part beyond the limits of MMB, you could even shave off some of the material such that it now fits in the profile and whatever other requirements are imposed could possibly be satisfied.

Did I say that right?

 

[3DDave]

In 2025, the exam is for 2009?

Anyway, not that there are common examples of practical or actual uses of this series of "for completeness" tertiary datum solutions**, the most that can be said of Fig. 7-36 is that the theoretical allowance is for a tangential extended plane that is farther or nearer to the axis of rotation without the planar feature allowing a normal projection of the axis onto it. The greater the distance from axis to tangential extended plane the greater the amount of material in the part.

A different example would be that the plane used as a secondary would be a small flat on the bottom of a round feature, allowing a direct normal projection of the axis on it, which would clearly play havoc with all these interpretations, rather than the example planar segments that are in the Goldilocks zone of not being too close or too far. The examples are also ignoring that rotation of the part to meet the datum reference frame is going to cause interference with the vertical face, meaning a practical datum feature simulator/TGC has to also have a translation capacity in the non-LMB cases.

The practical item would be what the mating part would likely have also to do. This could be avoided by a notch into the horizontal surface that allows that rotation without interference, but one would know to do that if there had been a practical example to drive this set of rules.

**Funny name for a secondary datum reference, but the first reference constrains 4 necessary degrees of freedom, leaving only one to consider. Axial displacement is left free, but plays no possible part in these examples. In any case, the problem is what to do when an additional restraint is required that would otherwise conflict with a higher precedence restraint. The normal tertiary problem involves a face that is parallel to the example viewpoint as the primary, the OD as secondary, and the small flat as tertiary.