QC inspector - pre-qual job test question

[gabimo]

ASME Y14.5-2009.
Square block dimensioned with +\-. ( height/width/length). Perpendicularity of the side feartures relative to the bottom surface is limited by:
a.) title block tolerance (assuming one is available)
b.) rule #1
c.) answer a or b, whichever is greater
d.) answer a or b, whichever is smaller
e.) not defined ( more information needed).


Same question as above, but drawing implies ISO GPS with NO symbol E attached to the size dimension.

 

[3DDave]

All 3 edges generated by these surfaces can be perfectly perpendicular to each other, however the faces may not be at right angles.

I'm not sure how this can be true. If all 3 edges are mutually perpendicular then that controls 6 degrees of orientation. 3 of those are redundant because each edge lies on 2 faces, leaving the remaining 3 degrees applied to making the faces perpendicular to each other. I'm sure there is a more rigorous proof.

I don't see how the interpretation of title block angle tolerances applied to implied right angles is vague. The general angle tolerance applies to all untoleranced angles and anything that appears to be at right angles is assumed to be 90 degrees. Since they are not explicitly dimensioned these right angles are not explicitly toleranced.

So, every pair of nominally perpendicular faces and every axis that is nominally perpendicular to any set of faces and vice versa is a candidate pair to be inspected. It's no wonder inspectors ignore the requirement.

 

[CheckerHater]

@pmarc:
Size does not create exception to default requirement, because it doesn't contradict the default requirement. Although I really stop caring at this point.


"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[pmarc]

CH,
It is a pitty that you stop caring, because in my opinion knowing how to deal with UOS type of notes is really important - not only for this particular thread, but in general.

We will stop here, if you like, but the last thing I would like to understand is why you think that the size + Rule #1 combo, which creates indirect 0.2 parallelism requirement, does not contradict the default 0.05 parallelism note. Is it because you think that the default parallelism note is actually a refinement of 0.2, and as such should be met addtionally to the size/Rule #1 requirement?

3DDave,
Please take a look at the attached picture. This is my attempt to visualize what I meant in my previous post.

http://files.engineering.com/getfil...-eebbad810b22&file=block_edge_face_angles.PNG

 

[CheckerHater]

OK, follow me here.
If I have Size tolerance .10 and parallelism of .05 represented by FCF, does size tolerance override FCF?
If I apply parallelism FCFs to everything that nominally appears parallel, will they be overriden by sizes?
If I replace several FCFs with one single note, what does it change?


"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[3DDave]

pamrc,

The diagram is showing as-manufactured variation. It does not relieve the inspector of measuring the angle across the edges as described. The measured angle will vary just the same as any other raw measurement.

 

[pmarc]

If I have Size tolerance .10 and parallelism of .05 represented by FCF, does size tolerance override FCF?

No, it does not.

If I apply parallelism FCFs to everything that nominally appears parallel, will they be overriden by sizes?

No, they will not.

If I replace several FCFs with one single note, what does it change?

If this single note does not use "UNLESS OTHERWISE SPECIFIED" (or similar wording), it changes nothing. But if "UNLESS OTHERWISE SPECIFIED" is used in that note, that changes everything, as I tried to explain above.

Oh, by the way. You asked me for a reference to support my point. I gave you one that in my opinion does the job. Could you in return offer a reference to support your point?

-------

The diagram is showing as-manufactured variation. It does not relieve the inspector of measuring the angle across the edges as described. The measured angle will vary just the same as any other raw measurement.

The point of the diagram was not to say that the inspector is relieved of measuring the angles across the edges. It was to show that measuring across the edges may in some cases give false impression about actual geometry (in this case perpendicularity) of as-produced faces. Another, and probably more important, point was to re-emphasize that relying on title block angle tolerances may lead to serious product quality issues. If there is no clear agreement on what to measure (and I think our conversation is a good evidence of that), this is well enough to say that during inspection the same issue, if not bigger, will happen.

 

[3DDave]

pmarc - what false impression? So far you have shown nothing to disagree about. If you had written that relying on Y14.5 implied 90 degree angles was a problem, then that is a problem, but also no disagreement on it being a problem. But tolerances? Not a problem at all. However, fixing the implied 90 means fixing it for all such angles, whether they are used by FCF controlled tolerances or not. Feel free to add all the explicit 90 degree dimensions you want.

 

[CheckerHater]

@pmarc:

Your quote doesn't say anything. It replaces one term without definition with another term without definition.
I am tired of arguing about the obvious with people in denial.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[pmarc]

I am tired of arguing about the obvious with people in denial.

Allow me to leave the "people in denial" part without a comment.

I just hope that this discussion showed at least a bit that an obvious thing for one person does not have to obvious for someone else. And moreover, that the someone else's point can be valid unless proven otherwise.

------
3DDave,
I am glad you think that so far I have shown nothing to disagree about. I am not sure what else I could say to better describe what I am up to here without repeating myself.

To close this topic, I would like to emphasize my firm belief that using title block angle tolerances in a form of "ANGLES ±30 MINUTES" notation (or similar) is in most cases like asking for troubles. For implied angles it is mainly because it is unclear as to which angular relationships they should be applied to and in the end in reality nothing is measured (I think here I could agree that this is rather the implied dimension concept's "fault"). For explicit angles without individual tolerance indication it is because of inherent ambiguity of the directly toleranced angle dimension. The way to eliminate this ambiguity to some extent would be to use origin symbol at one end of angular dimension, but how many drawings with origin symbols applied to angle dimensions have you seen? I have seen very few.

 

[greenimi]

pmarc,

Re:"If this single note does not use "UNLESS OTHERWISE SPECIFIED" (or similar wording), it changes nothing. But if "UNLESS OTHERWISE SPECIFIED" is used in that note, that changes everything, as I tried to explain above."

On my replay on June 10, I did not use the verbiage;"UNLESS OTHERWISE SPECIFIED"

My statement--copy-paste:
"Therefore, considering a title block defined like:
.X ±.05
.XX ±.01
.XXX ±.005
.XXXX ±.0015
ANGLES ±30 MINUTES

INTERPRET DIM. AND TOL. PER ASME Y14.5-2009/ ISO GPS --as applicable "

Am I a little bit safer from "multiple legal interpretations"? Just curious.

And by the way, how would YOU answer the OP question and also my "extended" question if no UOS note is used?
If you prefer to not give me an answer (and the answer would be read as "in/with all practical purposes) I am perfectly fine with that. I do understand your point (about UOS) and I think it is a valid one and you don't want to promote in any size, shape or form incorrect/not valid material/advice on this forum. I perfectly understand that if only one time you are not consistent/consequent the horse will jump the fence.

 

[CheckerHater]

@pmarc:

You argue that dimension can override orientation control due to Rule 1.

You know better than me that RULE 1 CAN NEVER CONTROL RELATIONSHIP BETWEEN TWO FEATURES, EVER!

You chose to ignore the most fundamental rule of GD&T just to settle personal score; this is what I call "denial". I will gladly accept better term if you come up with one.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[pmarc]

On 7 Jun 16 greenimi asked:

And since I am all for learning and developing technical abilities, I would extend/change a bit the above two questions (just for fun) and let’s say if the question (s) were about parallelism (and not about perpendicularity) how the answer(s) will change (if even will)?

Your answer was following:

In ASME the answer would be "d."

The answer "d" in the original question is following:

d) answer a or b, whichever is smaller

where:
a,) title block tolerance (assuming one is available)
b.) rule #1

So on June 7th you claimed that rule #1 might control features parallelism (if smaller than title block tolerance), but today you are saying:

RULE 1 CAN NEVER CONTROL RELATIONSHIP BETWEEN TWO FEATURES, EVER!

May I ask what changed in between? Or you know what... I do not really want to know, because I have a feeling that you will come up with another interesting explanation that definitely will not bring us any closer to solve the UOS note dillemma.

Nevertheless, thank you for this interesting and (still) funny dispute.

 

[pmarc]

greenimi,
Please allow me to leave your question unanswered.
I already regret that I jumped into this discussion with my side questions.

 

[CheckerHater]

@pmarc:

I am always having hard time to distinguish if people are honestly misunderstanding, or intentionally distorting the truth. But I'll bite.

This is another thing that you know better than me.

Rule 1 is INDIRECT control acting when any other control is ABSENT.

You argue that Rule 1 is overriding DIRECT control that is actually PRESENT.

Do you honestly cannot see the difference between the two? I have to agree with you - it's really funny.



"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[greenimi]

I just want to say that I feel guilty that I put you guys thru this. (in essence, the OP asked a question and I extended the OP’s dilemma to parallelism which parallelism created the whole conundrum). Honestly, I had no idea will create such of issue. I self admitted that the additional question was for my own education. I am very sorry and I will get the “closed mind of the month reward”. And all of this because ASME rule#1 default. You see now, why I am not in love with ASME. But as I stated before I don’t like ISO either.
One is not better than the other, they are just different.

Therefore, in my opinion, discussions as such, will help many, many, many people to understand the fine details of the drawing dimensioning in general and relationship and/or agreement/conflict of the standard with their own imposed title block tolerances. So, no, not ever regret you guys jumped with both feet into this discussion. I think you helped tremendous number of otherwise “unaware” and “silent” (for this discussion point of view) people. In fact, that is why you fine gentlemen came here on this forum in the first place, don’t you? You guys fulfill the purpose on why this forum exist in the first place: To help people. And you did. THANK YOU!

 

[Belanger]

No worries, greenimi. I'm surprised none of the exchanges used emoticons. There are some funny ones on here that might have been appropriate!

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

 

[3DDave]

I just read a discussion on whether the C programming language should have specified the number of bits in an integer. This has been a walk in the park in comparison. No emoticons their either, but it's a developer blog MSDN blog.

We didn't even touch on the fact that if the section is rhomboid, there are unopposed surfaces. Sigh. Those are always the best red herrings. Fresh canned.

 

[mkcski]

To all who posted:

I am relatively new to the forum world. I must say I am completely blown-away with how the discussion progressed and how it "ended", especially greenimi's post. Real professionalism here!!!

I would like offer a non-technical observation: The disagreements as to how to apply block tolerances and their questionable interpretation promote nothing but non-productive, profit robbing time. I have seen way too many arguments on the shop floor where Mfg made a "good" part and QA labeled it "bad" based solely on the application of block tolerances. They then call Engineering to resolve the conflict who has no clue as to what the limits of functional acceptably are - the Engineer copped-out and just defaulted to the title block. Meanwhile the machine tool, at $150 / hour, sits idle while the three posture. Had the design been "studied" by Engineering to define the limits, and had GDT been applied to communicate them, the "argument" would never have occurred and the machine tool would have kept running and the part delivered on time.

Continuing: this post again confirms my belief that tile block tolerances, if used at all, must be limited to the size tolerance on features-of-size and should never (attempt to) define default relationships between features in any way. Meaningful relationships between features require specified datums (reference frames) and, in my opinion, due to the almost unlimited possible relationships, it is not realistic to define default datums in a title block or any other document. Any relationships between features must be specified on the field of the drawing with the appropriate controls.

 

[3DDave]

mkcski,

The problem to gaining agreement is that not all of the people in that discussion have the background to make the decision.

What is most telling is that QA should have been involved in the selection of a process to make an acceptable part before the machinist even got the task. The machinist should be able to produce a history for similar parts or similar features as a guide to what they can do, to QA so they know ahead of time if the part will likely be acceptable within a good margin. And QA should discuss the proposed measurement setup with Engineering to see if it is going to be performed as expected.

Why not involve Engineering with the manufacturing discussion straight-away? Because the drawing is of an acceptable part, not a manufacturing drawing and QA is charged with accepting or rejecting the parts. If the part is made out-of-house, there may not be any manufacturing to have a discussion with.

Neither QA or manufacturing should have any say about what the part does, unless the entire system is being analyzed, as sometimes small tolerance for variation in one part is to allow more tolerance for variation in a mating part or to remove assembly variation that installation techs would have to make up.

Applying FCFs (plus-minus -is- GDT) is no guarantee. I have discussed with an inspector who believed that a chained set of basic dimensions meant that the hole (in this case) was to be measured for location from the nearest chained feature, and not the datum in the FCF, because "that's what the engineer meant." I asked if he'd called the engineer, but this guy was a mind reader and didn't care to make that call.

So far I have not seen any example of a drawing where it is unclear what should be inspected simply because of title block tolerances. It doesn't matter if the tolerance is co-located or in a note or in the title block as long as there is a method to determine which is the correct tolerance to apply.

My suggestion is not that title block tolerances are good or bad, but they have exactly the same flaws as applying the same tolerances to explicitly applied linear and angular dimensions. And recall, where a 90 plus-minus assumed angle relation is replaced by a 90 degree basic relation when using an FCF, the problem of determining where the 90 degree angles are is not entirely eliminated.

 

[mkcski]

3DDave:

I was hoping for other perspectives on title blocks. Much appreciated. I don't have time right now to compose a response but I intend to in the near future.