Conical datum features/ Linear Extruded Shape/ Complex Shape 2

[greenimi]

Can a primary datum feature such as Conical datum features/ Linear Extruded Shape/ Complex Shape be specified at MMB?

In other words, can a cone be specified primary datum feature with an MMB modifier?
4.17-2009 states that the “irregular features of size” can be specified at MMB, LMB and RMB, but does not specify anything about any conditions (such as being secondary or tertiary ---not primary---) in order to be called at MMB and all the examples shown (unless I am missing something, which is entirely possible) that have the irregular features of size at MMB/ LMB are depicted with this datum feature being secondary and tertiary (not primary at MMB)

Any condition, that you are aware of, is applicable here?

I know the standard cannot cover every possible scenario, but I am wondering, if something is “forbidden” by other rules (and regulations) in the standard that I am not aware of.

 

[powerhound]

You can do it but since it is a primary datum reference it will have absolutely no effect. You may as well just leave it RMB. What is it that you think you can do at MMB that you don't think you can do at RMB? Maybe there's another way.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[drawoh]

greenimi,

How are you going to fixture a primary-datum cone? You have size and angle to account for.

--
JHG

 

[3DDave]

This is one of those areas where the standard is vague for conical and complexly shaped items as datums. I don't recall that 14.5.1 is any better.

A cone doesn't have a simple evaluation for size measurement and it doesn't have a simple locating or orientation evaluation when it is imperfect. If a cone is not significantly deformed, a conical fit resolves to a point and an axis, neither of which has a size, so in practice I would avoid any size related modifiers. As discussed before, my feeling is the 'size' of a cone, as depicted in the standard, is a side-effect of allowing variation in the location where a single diameter measurement is made. In order to 'see' that size, the cone has to be oriented and located in a particular frame of reference, so it has no unrelated actual mating envelope. Whatever value is desired can be found by moving to a different location on the body of the cone indicating that cones are always the desired size.

 

[drawoh]

3DDave,

I am still trying to visualize a fixture. Forget about primary datum. The primary datum feature is the large, flat end opposite the point. The secondary, locating datum is where a Ø50mm ring with a 2mm[ ]radius fillet contacts the cone. Add a clocking feature if necessary. If the ring has a fixed height, you have sort of a functional MMB. The cone can have all sorts of weird configuration away from the Ø50mm[ ]ring.

--
JHG

 

[greenimi]

drawoh,

We have a design similar (or almost similar) with the attached picture I found in one of the GD&T books and also in their resource online. Other datum feature called B (secondary) is used -to stop the remaining roation degree of freedom , but suffice to say that A is primary --as shown--
Someone questioned if A (primary) can be (legally) modified at MMB.

Also, the extended question was if similar approach can be used for Linear Extruded Shape and Complex Shape--kind of shapes shown in Fig. 4-3 ( f and g).

 

[3DDave]

drawoh,

When I replied to begin with there were no other posts.

There isn't a way to inspect the cone 'size' without relying on pre-defining another feature as an orientation and location reference, so it hasn't got a size to refer to. A datum target on the cone? OK, but that's not a size of the cone.

 

[Nescius]

Why must a datum feature simulator for a conical datum feature have the capacity to change angle? I would argue that the simulator should NOT adapt to the angle of the feature.

Edited to add the attached PDF. Of course, in depicted cases A and B and C, nobody would argue that the simulator for datum feature B should be anything other than fixed as perpendicular to the simulator for datum feature A, regardless of how irregular the actual feature may be produced. Then in case D, why would the simulator suddenly need to be adjustable in angle? It would just be a conical surface, 179.8 degrees included angle, fixed, and perfectly oriented to the simulator for datum feature A.

It's just a revolved surface. Simulators for inclined planar datum features do not adjust in angle, so why would a conical simulator behave in this manner?

 

[Nescius]

Having trouble attaching my PDF in an edit. Let's try here...

 

[drawoh]

Nescius,

A cylinder locates in X and Y. If (when) the cylinder is below perfect form at MMB, the result is that it can move in X and Y. Z is controlled by some other feature. A cone locates in X, Y and Z. At anything other than MMB, how should your fixture account for an incorrect angle?

I have no issues with a cone as a secondary datum. It provides X and Y location. You can specify positional tolerances at secondary MMB. You can implement RFS by making your secondary fixture variable height.

--
JHG

 

[Nescius]

An imperfect conical primary datum feature referenced at RFS might very well rock in the fixture. Of course, an imperfect cylindrical primary datum feature referenced at RFS could also rock, say, if the feature is tapered.

I am not sure that referencing a conical primary datum feature at MMB has a real meaning.

 

[greenimi]

I am not sure that referencing a conical primary datum feature at MMB has a real meaning

Exactly.... that is my question, too.
Probabbly is legal, but not usefull.
So, final answer?

Anyone else can provide some input based on their experience-application based? Or even some paragraphs/ verses from the standard that I am missing.

Thank you again for your help

 

[greenimi]

And again, the question (s) is/are not neccessary for cones olny, but for Linear Extruded Shape and Complex Shape--kind of shapes shown in Fig. 4-3 ( f and g) TOO.

 

[Belanger]

...Or even some paragraphs/ verses from the standard that I am missing.

Not that this relates directly to the cone idea, but since you asked for stuff from the standard: check out Figure 4-21(c) for an example of a primary datum at MMB.

So just to be clear -- it's not necessarily a problem to have MMB on a primary datum (with a secondary datum still referenced).

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[3DDave]

Figure 4-21 (b) and (c) demonstrate that the secondary reference is not required in that example. A DOF analysis shows that the plane in [A(M)|B] doesn't control any independent DOF(s) for the 4 holes. Particularly (b) shows a lower priority datum reference gaining orientation control over higher priority reference in an ad hoc fashion. I see it as an example that should be clearly labeled "Not Recommended" along with an explanation.

Compare '2009 4-20 and 4-21 to '1994 Figure 4-18.

 

[powerhound]

it's not necessarily a problem to have MMB on a primary datum

I was referring to the OPs question regarding a conical feature as primary. I wasn't clear that I was only referring to the conical feature in question but I would have said the same thing about a plane as primary at MMB. You can do it but it has no effect. In the case of 4-21 the effect is profound.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[pylfrm]

Can a primary datum feature such as Conical datum features/ Linear Extruded Shape/ Complex Shape be specified at MMB?

I will assume that the proposed primary datum feature is fully defined with basic dimensions and toleranced with a single uniform equally disposed bilateral profile of a surface tolerance. If anyone has something else in mind, please say so.

A linear extruded shape (such as Fig. 4-3, illustration (f) of ASME Y14.5-2009) is an irregular feature of size. I see no problem with referencing it as primary datum feature at MMB. I'd say the datum feature simulator is the maximum material boundary of the profile tolerance, and the part is free to shift around within (and optionally make contact with) that boundary.

Planar, conical, and complex features (such as Fig. 4-3, illustrations (a), (e), and (g) of ASME Y14.5-2009) are not features of size. I think the standard is much less clear in this case, so I would try to avoid the issue where possible. However, you might want to look at para. 4.13 regarding mathematically defined surfaces.

We have a design similar (or almost similar) with the attached picture I found in one of the GD&T books and also in their resource online. Other datum feature called B (secondary) is used -to stop the remaining roation degree of freedom , but suffice to say that A is primary --as shown--
Someone questioned if A (primary) can be (legally) modified at MMB.

Why did they ask? What were they hoping or expecting to accomplish with the change? Also, could you describe datum feature B (and any others involved) on this part in more detail?

I was only referring to the conical feature in question but I would have said the same thing about a plane as primary at MMB. You can do it but it has no effect.

I would be very hesitant to say this. If you accept that it's valid at all, I think it's hard to argue that the datum feature simulator shouldn't be shifted away from the true profile to the maximum material boundary. I'd be interested to hear your thoughts on this.


pylfrm

 

[axym]

Hi All,

I agree with everything that pylfrm said, and I'll add a few additional thoughts.

The questions in greenimi's original post are difficult to answer, because they probe into areas that Y14.5 is vague on.

Y14.5's RMB and MMB concepts were built around regular features of size, for which the constraint effects of material condition are fairly clear. They are difficult to generalize to more complex geometry, especially shapes with special properties such as cones and wedges.

There are some clues in Y14.5-2009's section 4.16 on Rotational Constraint About A Datum Axis or Point. If we go through figures 4-29 through 4-32, some general principles are apparent:
-RMB means that the datum feature simulator does not translate or rotate, but progresses to make maximum contact with the datum feature
-MMB means that the datum feature simulator is fixed at the MMB and does not translate or rotate, or progress. Fig. 4-31 implies that the datum feature must make contact with the simulator, but this is not stated anywhere.
-[BSC] means that the datum feature simulator is fixed at the basic profile and does not translate, rotate, or progress. The datum feature does not need to make contact with the simulator if the rotational degree of freedom is restricted in both directions (as in Figs 4-29 and 4-30) but needs to make contact if the rotation is not restricted (as in 4-31). I've never been comfortable with this inconsistency.

There are also some clues in section 4.13 Mathematically Defined Surface. The primary datum feature in Fig 4-28 is a mathematically defined surface, but it's practically planar. The figure shows it referenced at [BSC], and the text states that the high points of the datum feature are aligned with its simulator. The text also talks about what would happen if the datum feature was applied at MMB or LMB, clearly indicating that these options are possible and legal. Whether these options make sense or not is another question entirely, of course.

Now we can go back to Fig 4-3 and assess which material condition modifiers (RMB, MMB, LMB, [BSC]) could be applied to each feature type as a primary datum feature. Here's what I would get:

(a) Planar. Only [BSC] makes sense to me. MMB and LMB are possible, but wouldn't make sense because the simulator (and thus the datum) would be shifted relative to the true profile. RMB would not work, because the simulator would have no curvature and maximum contact would not be definable.

(b) Width. This is a regular feature of size, so RMB, MMB, and LMB are well defined and make sense. [BSC] is possible but would not make sense, because the datum feature may not fit over the simulator.

(c) Spherical. Regular feature of size, so same as (b).

(d) Cylindrical. Regular feature of size, so same as (b).

(e) Conical. To me, this is similar to the planar feature type. Only [BSC] makes sense to me. MMB and LMB are possible, but wouldn't make sense because the simulator would be shifted axially relative to the true profile. RMB would not work, because the shape of the simulator doesn't change as it offsets (special property of cone). Maximum contact would not be definable.

(f) Linear Extruded Shape. To me, this is similar to the regular features of size. RMB, MMB, and LMB are well defined and make sense. [BSC] is possible but would not make sense, because the datum feature may not fit over the simulator. One additional detail here is that it's assumed that the linear extruded shape is "closed" like the one illustrated, such that the datum feature envelops or is enveloped by the simulator. If the shape were "open", then things might be different.

(g) Complex. Anything goes - RMB, MMB, LMB or [BSC] could all be applied. But it would be case specific - which ones make sense would depend on the particular geometry of the complex surface. The quasi-conical complex surface in Fig 4-3 (g) is very different from the quasi-planar complex surface in Fig 4-28.



Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Nescius]

Evan,

I get what you're saying, but I'd go further. I'd reject applying the concepts of BSC vs. RMB (I mistakenly said "RFS" in my previous post) to a planar primary datum feature. And cones, too, I think.

I think the discussion centers around this point: We don't reference a primary datum feature for its own sake, we reference it to act as a...reference. The ways we put "handles" on certain shapes does not compute with the rules of GD&T because some shapes don't have "size", or at least not in the traditional sense. In the cases of planes and cones, we might be able to imagine their own flavor of "size" as a variable thing, but not one that is independent of the "handles" we use to know those shapes.

what does the collective think about the attached drawing. Here, the primary datum feature is not a regular feature of size, but a "handle" we use to mathematically "touch" the datum feature is independent of the non-traditional notion of size that this shape has. I'd argue that any of the position FCF's could be properly confirmed with a simple fixture. The partial sphere is frustratingly similar to a blunt cone, but the "handle" of the cone is the point, and you can't have a "bigger" cone with the same point in the way that you can have a "bigger" partial sphere with the same center.

 

[drawoh]

Nescius,

I take it that that RXXX feature actually is SRXXX?

Is your piece solid, or a piece of spun sheet metal? This would affect how I would apply datums and tolerances.

--
JHG