Rotation 1

[aniiben]

ASME Y14.5-2009
Part assembled on A (flat surface) ( flatness considered--not shown in this sketch for simplicity) then
centered on B (OD) (perpendicularity to A considered--not shown in this sketch for simplicity) and
bolted down with 4 screws (Ø.475 holes) (basic dimensions considered--not shown in this sketch for simplicity)

The other 4 holes (not dimensioned for size in my sketch) are non-functional (aid-process only)

Not sure about rotation.

Can I use C(modified at M) to stop the rotation only? Nothing is stopping the rotation except 4 bolts in the same time.
Is A, B(M) and C(M) valid? Legal? Why? Why not?

 

[chez311]

semiond - I read and understood the functional requirements of the different features per the OP's notes, thats not what I was asking about.

What I was asking was whether the gauges and actual tolerance zones between the three schemes I mentioned were any different. From what I can tell the answer is no, which confused me which is why I asked the question.

The only scheme that describes the OP's design intent, of the 3 you mentioned, is the first one.
The second one can not guarantee the function for all possible tolerance combinations.

Per pmarc's reply "You are right, both [A|B(M)|C(M)] and [A|B(M)-C(M)] accomplish the same thing." so unless you are disagreeing with that, the only thing calling out one vs. the other does is communicate something different to whoever is reading the print - the actual result is the same.

The third one completely puts the hole pattern out of the game, unless the pattern itself is toleranced with the same datum call outs and therefore subjected to the simultaneous requirement.

In my hypothetical the pattern would be held to the same DRF as the rest of the features which would be [A|B(M)]. Per my above, as far as I can tell this accomplishes the same thing as the other two situations, but I would interested to be shown otherwise because it doesn't FEEL right intuitively but thats the way it seems to me.

 

[semiond]

chez311,
I agree the gage fixtures structure might be the same for all 3 cases. But, the design intent is different - and therefore (and here we get again to the point i was trying to convey to aniiben earlier in this thread), the specified tolerance values for the features should be different for each case, to be in accordance with each particular design intent. This might result in different SIZES of comparable features in the 3 different gage fixtures. Again, ideas on how to specify those values correctly you can find in pmarc's suggestions.

It is not only the datum reference frame that is specified in a drawing that makes the part function like it should. The actual tolerance values are even more important, i would say.

 

[chez311]

I agree the gage fixtures structure might be the same for all 3 cases

Looks like we agree then! Not to brush aside what you are saying about tolerance values and design intent as I think thats very important and relevant to what the OP is trying to accomplish with their drawing, but the core of my question was essentially if everything stayed the same between the three cases besides changing the DRF in each FCF would the resultant tolerance zones/gauge be any different for the three cases and you seem to have agreed with my analysis that they would the same.

I'd be interested to see if anyone else supports/refutes that.

Didn't mean to hijack this thread, as I do believe it is still in line with the original inquiry. Its just one additional step towards my having a better understanding of MMB - I hope it might help someone else too.

 

[semiond]

chez311,
I think i understand your current mindset.
You want to put the focus for a while only on the technical aspect of gaging and tolerance zones, and later make conclusions on other things from that.

That's fine, and i'm looking forward to other responses on that too. But let me say something that might be more in the spirit of your query:

1. I am almost sure that if everything stayed the same besides the DRF for the callouts, the same gage could be used to inspect the part in all 3 scenarios.

2. I think that for a given set of tolerance values, only one datum reference frame will always correctly describe the function of the part, which results from the above mentioned set of tolerance values.

3. I think that a combination of a datum reference frame and a set of tolerance values that do not describe the same function, is a bad design practice.

I find point no. 1 the least important, and point no.3 the most important.

 

[chez311]

semiond - you make several fair points, and I can appreciate your viewpoint. I certainly agree that it is absolutely crucial to set up a drawing that most clearly communicates the design requirements through the datum structure in combination with consideration of the tolerances that properly reflect those same design requirements.

That being said sometimes I feel it necessary strip concepts down to understand why certain things work the way they do - if my understanding of the core technical aspects of GDnT are lacking, its difficult to apply it properly at a higher level. In this case I was considering the difference between 3 cases/DRF's thinking that there was a difference in the actual resulting tolerance zones and difference in the effect on referenced features and realizing they all accomplished the same thing! (notwithstanding your noted objections about communication of design intent and associated tolerances)

 

[aniiben]

Should we conclude that the depicted case is legal, but not recommended and it is a poor design practice due to the “changed rules in 2009” from default orientation to default orientation and location?

Or this conclusion is a stretch and has been the same “poor practice” in 1994 either?

 

[drawoh]

aniiben,

I believe we have interpreted everything so far, as per the standards. How do you want your part fixtured for fabrication and inspection?

--
JHG

 

[aniiben]

I am wondering if there would be a different interpreation if 1994 standard is used versus 2009?

Yes, it would be inspected ....probably with a functional gage.

But at this point I am interested to find a correct and unambiguous part definition.

 

[3DDave]

The 1994 version included the fiction that the mating feature to the tertiary datum would translate in a radial fashion to the secondary datum and therefore only the width was important. Even with a diamond pin this is not correct. Ref Fig 4-9(b) in 1994, emphasizing the width of datum feature C.

In the 2009 version the default interpretation was changed to match realizable components and the 'translate' modifier was added to indicate the less realizable/fictional condition. Ref Fig 4-9(c) in 2009 which no longer shows the width and includes emphasis on the basic dimension locating the datum feature.

 

[chez311]

So after sitting on this for a bit I have an additional question on this topic - I hate to open what was a mostly closed topic however this has been bugging me.

pmarc - why are you concerned about the tertiary datum bolt holes controlling translation instead of rotation in this instance but perhaps not others? In any fixed gauge I would think this is a possibility. ANY feature on the part could constrain it in translation/rotation instead of the datum features (all called out at MMB in order to utilize fixed functional gauge) if the component is produced to certain limits - ie: in this case even if the 4x inner bolt holes are specified with undersized pins per your suggestion, now the 4x outer tighter tolerance holes, if produced near/at their MMC boundary and size, could constrain the part in translation instead of the central datum feature. In fact, I would have to see the size tolerance allowed on these 4x tighter tolerance holes as it is not provided, but assuming it is also tighter - it is likely these non-datum features could actually end up doing just this the majority of the time - especially if the 4x bolt holes are specified with undersized boundaries/pins per your suggestion!

I would think that the best way to combat this would be to specify your datums with tolerance values that match their precedence and are an order of magnitude more accurate than the rest of the features on the part to ensure that for the majority of the parts produced this does not happen and/or take special precautions where necessary (datums at RMB where applicable, diamond pins, etc..).

I hope my inquiry makes sense - let me know if figures would help.

 

[semiond]

I would think that the best way to combat this would be to specify your datums with tolerance values that match their precedence and are an order of magnitude more accurate than the rest of the features on the part to ensure that for the majority of the parts produced this does not happen and/or take special precautions where necessary (datums at RMB where applicable, diamond pins, etc..)

 

[pmarc]

chez311,
You make a good point and I fully understand your inquiry.

I would say that the key point here is one word that you used in this statement: "In any fixed gauge I would think this is a possibility".

I do not think that the gage has to be fixed. (And by that I do not mean that the size of the elements used to build that gage does not have to be fixed, but rather that the gage does not have to be a single body with all its elements fully fixed relative to each other).

There is a concept in gage design called sequential gaging. I do not know if you have access to ASME standard Y14.43-2011 "Gages and Fixtures", but in that document there is an example B-21 showing how the concept works in case of inspection of a part similar to the one shown in fig. 4-18 in Y14.5-2009.

It is hard to exactly describe the design of the gage (pictures would do much better job, but I am currently not able to attach any pictures), but basically the idea is to lay the inspected piece part on datum feature simulator A and then sequentially add remaining features to the gage - first datum feature simulators B and C, and finally the gage elements simulating virtual conditions of the 4 holes and the outer contour of the part.

This approach is different from much more common method of using a single-body gage by the fact that it allows to position/adjust the piece part relative to the datum feature simulators first (within allowable actual clearances), then hold the part, and then check if the gage elements used for verification of the toleranced features do not collide with the piece part. If they do in a first try, the part can be readjusted/repositioned relative to the datum feature simulators, and then the toleranced features can be verified again. As long as there is at least one part-to-gage configuration that allows all gage elements to fit in/over the actual toleranced features, the piece part is considered good.

So in the sequential gaging approach the datum feature simulators really can do their intended job.

 

[chez311]

pmarc - I only have access to an outdated version of Y14.43 whose appendix B does not have a figure B-21, it only goes up to B-20. That being said I can visualize what you are talking about, essentially allowing the datum features to do their job and shift around the datum simulators as gauge elements are added for each feature.

In this sequential gauging however, even after all elements are added I would think it can still result in a condition where the part is constrained by the aforementioned tertiary datum features or even other features on the part - right? However are you no longer concerned with this as it happened AFTER the datum features and then other features were verified in the proper sequence instead of during initial insertion of the part onto the gauge?

As an additional question - if the tertiary datum was changed to the tighter tolerance holes and the secondary datum was modified to be sufficiently more accurate than the tertiary datum (OR if the tertiary datum on a similar part was a single hole instead of a pattern of 4x holes) - would this no longer be as much of a concern and still be a good candidate for a fixed "single body" gauge?

 

[pmarc]

In this sequential gauging however, even after all elements are added I would think it can still result in a condition where the part is constrained by the aforementioned tertiary datum features or even other features on the part - right? However are you no longer concerned with this as it happened AFTER the datum features and then other features were verified in the proper sequence instead of during initial insertion of the part onto the gauge?

Yes, that is correct.

As an additional question - if the tertiary datum was changed to the tighter tolerance holes and the secondary datum was modified to be sufficiently more accurate than the tertiary datum (OR if the tertiary datum on a similar part was a single hole instead of a pattern of 4x holes) - would this no longer be as much of a concern and still be a good candidate for a fixed "single body" gauge?

It would definitely be better than the case with the tertiary datum feature toleranced with similar or tighter tolerance values compated to secondary datum feature.

 

[chez311]

Thanks pmarc - I appreciate your answering my questions, like I said I had been mulling it over for a bit and I figured it was time to put it to rest as I still had some open questions.

 

[pmarc]

You're welcome, chez311.

And I am actually very glad how this thread evolved. We touched one of the hottest (if not the hottest) GD&T topics on my list.

 

[chez311]

pmarc - I would fully agree, I've had a few important realizations while working through this.

Additionally, not to beat a dead horse but as an extension of this conversation I think I finally realize another issue arising when excessively sloppy features are utilized as datums from which tighter tolerance features are referenced. Obviously we've already discussed what happens at MMB (undesired location/constraint by other datums/features) however I am thinking at RMB potentially this could lead to issues with production as opposed to MMB where the datum feaures can float and shift vs. their simulators, the entire part must be inspected and fixtured directly to these sloppy datums (ie: with expanding collets/pins or similar adjustable fixturing) in order to produce a set of more accurate features.

Now a good production shop with proper fixturing may be able to handle this, however at the very least I imagine it makes everyones job harder than necessary.

 

[drawoh]

chez311,

This discussion is awkward because we do not understand what the part does. We have proposed tolerancing schemes that, however correct they may be, describe different parts. There was an excellent discussion about this very issue... thread1103-322065. The OP wanted to know how to identify critical dimensions that he had to inspect and meet. The implication of course, is that the other dimensions and tolerances called up, do not matter. This all is worked out in discussions with the designer, or with whoever assembles this thing. To heck with standards.

The OP has indicated in another discussion that his parts are secret or proprietary. This makes his understanding of functionality and of drafting standards that much more important.

--
JHG

 

[chez311]

drawoh - that was an interesting read on the topic of critical dimensions and the different viewpoints of how/when/if they should even be applied as well as the realities of inspection vs. design requirements. I agree with you that , as semiond pointed out, a particular datum and tolerance scheme should absolutely jive with the design requirements. That being said, while it was I think my questions were still on-topic with the OP's initial inquiry, I was sort of taking his initial example and applying it to a more general case of "what if's" which actually shed a little additional light on datum selection, MMB, and gauging for me. I hope it could do the same for someone else while maybe giving the OP a little more ammo with which to make a decision on datum/tolerance schemes.