Fig. 6-30 in ASME Y14.5-2018 - basic location of a thread start? 3

[Burunduk]

In fig. 6-30, what tolerance controls the length of the threaded M42 section?
There are basic dimensions 22 and 32, however while basic dimension 32 is clearly related to the general profile tolerance since there is a step there that is part of a true profile, I'm not sure what to make of basic dimension 22. It is given to where the basic diameter 36 ends and the thread starts, but is the true profile and the general profile tolerance zone well defined at that place? Part of the problem as I see it is that there is no step there to take part in a clear true profile (basic dia. 36 is only slightly smaller than the minor dia. of M42).

 

[Burunduk]

Even though both dimensions do not have standard intepretation in Y14.5, that's usually too weak argument to succesfully convince most people to change the scheme to basic dimensions and profile.

Unfortunately so.
Maybe with the industries moving to model based definition or minimally dimensioned drawings with a basic model, where not every basic dimension has to be graphically stated, it will become easier to stick with robust definitions without causing anyone to cringe.

 

[pmarc]

I agree that MBD should help, but only if organizations start to see value in seamless downstream consumption of the definition. I have seen quite many so called "MBD definitions" where not even 10% of the callouts would be consumable by a CMM, or tolerance analysis software. And it was not the fault of the CMM or the tolerance analysis software, but of the poorly prepared definition, full of directly toleranced non-size dimensions, directly toleranced angles, missing associated objects in the CAD model, etc., prepared to simply mimic "the good old fashioned 2D definition that has worked for 30-40 years, therefore not requiring any improvements".

 

[Burunduk]

pmarc,
Very well said. I wish ASME Y14.41 didn't support directly toleranced angles. Do you think they will ever be gone?

 

[greenimi]

I agree that MBD should help, but only if organizations start to see value in seamless downstream consumption of the definition. I have seen quite many so called "MBD definitions" where not even 10% of the callouts would be consumable by a CMM, or tolerance analysis software. And it was not the fault of the CMM or the tolerance analysis software, but of the poorly prepared definition, full of directly toleranced non-size dimensions, directly toleranced angles, missing associated objects in the CAD model, etc., prepared to simply mimic "the good old fashioned 2D definition that has worked for 30-40 years, therefore not requiring any improvements".

There are a couple of impediments for the MBD journey. One that comes in my mind is: how do the MBD practitioners know which annotation is machine-readeable and which annotation is only human readeable. The same way which dimension/ or annotation is CMM readeable (or QIF) or which one needs humans to interviene. Or at least that is my problem which I am not sure how to solve.

Another issue is unclear and very muddy definition of the UOS. ISO GPS did a very good job in ISO22081, but on the ASME side we have to do some "nasty adjustments". At this point, I am not sure when the general profile is applicable and when is not, in order to be sementically associated with the feature.

pmarc,
Do you have experience in this area of experise?

 

[pmarc]

I wish ASME Y14.41 didn't support directly toleranced angles. Do you think they will ever be gone?

Burunduk -- I really don't know. I assume (but that's just assumption) that in case of directly toleranced angles and a few other commonly accepted applications of directly toleranced dimensions that are mentioned in Section 10 of Y14.41-2019, they won't step ahead of 14.5. Although they probably should, because their input from the view point of automated downstream digital consumption of the definition could be truly beneficial, just like in case of the (U) modifier that had been first introduced to Y14.41-2003 before Y14.5 implemented it in 2009.

There are a couple of impediments for the MBD journey. One that comes in my mind is: how do the MBD practitioners know which annotation is machine-readeable and which annotation is only human readeable. The same way which dimension/ or annotation is CMM readeable (or QIF) or which one needs humans to interviene. Or at least that is my problem which I am not sure how to solve.

Another issue is unclear and very muddy definition of the UOS. ISO GPS did a very good job in ISO22081, but on the ASME side we have to do some "nasty adjustments". At this point, I am not sure when the general profile is applicable and when is not, in order to be sementically associated with the feature.

pmarc,
Do you have experience in this area of experise?

greenimi -- Regarding readability, I like to think about it this way. Machines (CMM's, CNC's, tolerance analysis software, etc.) are programmed by humans that use (or at least should use) standards like Y14.5, Y14.5.1 or Y14.41 to know how to do correct programming. If the standards do not provide clear interpretation for a tolerancing tool, the humans will not know what to do and the machines will not be able to read that.

As for the UOS issue, I agree that ISO did it quite wisely and that in ASME there is still some work to do. I do believe, however, that in MBD this problem can be mitigated to some extent by associating appropriate features of the model with the UOS callout in the CAD software used to prepare the definition. In other words, those features that have not been associated with the UOS callout in the CAD model are simply not controlled by it. Of course, this may work only if the organization and their business partners (customers, suppliers) have tools to be able to query the definition for the associated features.

 

[ewh]

In a previous position creating full MBD models, one of the primary rules was that ALL features that didn't have a specific tolerance or FCF associated had to be associated to the default FCF. Every surface of the model had to be associated to a tolerance of some sort.

"Know the rules well, so you can break them effectively."
-Dalai Lama XIV

 

[greenimi]

In a previous position creating full MBD models, one of the primary rules was that ALL features that didn't have a specific tolerance or FCF associated had to be associated to the default FCF. Every surface of the model had to be associated to a tolerance of some sort.

So if, for example, the designer "forgot" or missed the circular runout in Burunduk's first picture (from 2018 standard) for the 20.00-20.13 inside diameter feature or for 31.6-31.8 outside diameter feature, would you still associate this feature with the general profile?

Per the statement in the above quote, the features ( those two exemplifying above) did not have FCF associated.
Then what to do? How to read the MBD model?

Pmarc (since I do know your level of expertise and experience), how would you recommend to solve this issue? (Although very often encountered on the drawing level)

Many thanks

 

[Burunduk]

So if, for example, the designer "forgot" or missed the circular runout in Burunduk's first picture (from 2018 standard) for the 20.00-20.13 inside diameter feature or for 31.6-31.8 outside diameter feature, would you still associate this feature with the general profile?

Greenimi, the features you mentioned are specified with a direct size tolerance. In this case, the model value can not be considered basic. If there is no basic shaft diameter, there is no unambiguous true profile and the general profile tolerance doesn't work.
If someone "forgot" to specify runout or position tolerances for those directly toleranced diameters, they should be contacted and told there's a location control missing.

 

[greenimi]

Burunduk,
Based on your latest replay, you would consider the MBD incomplete, correct?

 

[Burunduk]

Yes, it would be incomplete.

 

[pmarc]

greeinimi,

A full MBD definition means that all features of the model have all necessary tolerances applied and that the tolerances have been associated correctly with the features. So if a surface of the model, like the diameter 20.00-20.13, has only the size requirement associated with it, then that surface is not controlled for orientation and location at all, despite that the general profile tolerance has also been used. Moreover, even with the runout tolerance specified, the surface may still not be fully defined in MBD sense because someone might forgot to correctly associate the runout tolerance with the surface in the CAD model - this of course will be interpretable by humans, but the machines will have no way to know that the runout tolerance applies to this particular surface.

Technically speaking, the UOS term is redundant (or maybe I should say auxiliary) in the full MBD approach, because querying the requirement specified at the UOS basis should give a reader necessary information as to which features of the model are controlled by it, therefore there is no need to say that the requirement is UOS.

 

[Burunduk]

Technically speaking, the UOS term is redundant (or maybe I should say auxiliary) in the full MBD approach, because querying the requirement specified at the UOS basis should give a reader necessary information as to which features of the model are controlled by it, therefore there is no need to say that the requirement is UOS.

I second that. There isn't really a need for a UOS tolerance in MBD. One can simply specify a loose profile tolerance to the appropriate group of surfaces (with the associativity as required by Y14.41 of course). UOS makes sense with minimally dimensioned drawing method when the model is not fully annotated.

 

[greenimi]

Technically speaking, the UOS term is redundant (or maybe I should say auxiliary) in the full MBD approach, because querying the requirement specified at the UOS basis should give a reader necessary information as to which features of the model are controlled by it, therefore there is no need to say that the requirement is UOS.

Pmarc and Burunduk,
Should I understand that you ARE saying that you agree with a note such as "ALL UNTOLERANCED SURFACES OR FEATURES: PROFILE|XXX|DRF| as long as the wording "UOS" (Unless Otherwsie Specified) is NOT used.
Did I understand you correctly?
I am wondering about this group of words (UOS) and what could be so special about them and why "they" are so dangerous.

 

[pmarc]

greenimi,

What makes you think that I would agree with the note you proposed? There should never be such thing like UNTOLERANCED SURFACES OR FEATURES. Each feature should have a tolerance.

The point I was trying to make is that in MBD one can theoretically replace the UOS profile callout with just a profile callout and upon querying the callout, it will become clear which features it applies to.

 

[Burunduk]

greenimi, I suppose that with "UNTOLERANCED SURFACES OR FEATURES" you try to borrow from the old-fashioned general tolerance wording "UNTOLERANCED DIMENSIONS..." but really there should be no untoleranced features, as pmarc noted. The idea is that a proper MBD model should include associativity between the tolerance, the considered features, and the datum reference frame (where applicable), for every requirement. In that sense, the "general" profile tolerance defined on the model should not behave differently than any other feature control frame on the same model (even if it lacks a leader directed to any model surface). Hence words like "UNLESS OTHERWISE SPECIFIED" don't say anything relevant (for MBD).

 

[greenimi]

I suppose that with "UNTOLERANCED SURFACES OR FEATURES" you try to borrow from the old-fashioned general tolerance wording "UNTOLERANCED DIMENSIONS..." but really there should be no untoleranced features, as pmarc noted.

You are correct.
I am trying to convert everything I see on the drawing to the model. And since general profile is shown on the drawing then, consequently, I think the UOS should be included in the mix. However, looks like you guys disagree.
Do you think, the scope of general profile and the UOS note is achieved if the model is left incomplete?

By the way, what is the main scope of the UOS note: to reduce the drafting annotations (keep the drawing less crowded) or to complete (granted artificially) the feature's definition of the part?
Or maybe is neither one?

What is your opinion?

There are people, seen opinions on different forums, that the scope of UOS note is to make the drawing complete.
(Something like the major benefit that general geometrical tolerance is to make the drawing complete, instead, to potentially leave any tolerance open for users to define). What do you guys think?

 

[Burunduk]

greenimi, let's start by this:
If the product definition follows the rules and the supported practices, whether it is drawing based or model based, there may be untoleranced dimensions (basic or reference), but not untoleranced features.

Now, with that in mind, I think the answer to your question about the UOS note (or the general profile tolerance) is that its scope is primarily to complete the definition of the geometrical variation limits for the non-critical features of the part. BTW,Why "artificially:?

 

[greenimi]

BTW,Why "artificially:?

Because you rely on a note (general note) and not on a well thought and functional need to fully define your part.

 

[Burunduk]

greenimi, just because it's a general note, doesn't mean that its use shouldn't be "well thought and functional". Deciding which are the more functionally important features that get unique tolerances defined for them and which are the less critical features that can be controlled by the general note takes some consideration and forming of a clear design intent. Assigning the appropriate profile tolerance value also requires consideration.

 

[pmarc]

Well said, Burunduk.

I would just like to add that although it is common practice to specify general profile tolerance with the value that is the most generous of all tolerance values used on the definition (i.e., for the least important features from product function point of view), this is not a requirement.