After titans ôfightö user left confusedùôback-door locationö

[greenimi]

After the thread which debated the “back-door location” I am a little confused on how should I read the following drawing-see attached sketch—

I have created 3 cases to support the conversation (some drawings incomplete by intent)


Case #1: has 2x .175/155 dimension and parallelism callout to datum A only (1.092 is reference).
Is this drawing fully defined? In my opinion yes



Case #2: no 2x .175/155 dimension, but has parallelism to datum A and B.
(1.092 is not reference)
Is this drawing fully defined? As per the thread, I would say yes.


Case #3: no 2x.175/155 dimension and parallelism to datum A only.
(1.092 is not reference and centerline shown for .769/.755 dimension)
Is this drawing fully defined? I don’t know.



Thank you again for helping “little” guys like me and supporting a learning environment.

 

[Belanger]

Dean, I believe that the 1994 standard didn't even have an equivalent to paragraph 4.5.2 of 2009. (Don't have the old std in front of me right now, but please check.) So it's not just the translation idea that they added there, but also explicit mention of how the simulators must be related to each other.

IOW, nowhere did it say that the simulators must be basically located from each other, which opened the door to discuss an offset simulator for the slot in Frank's earlier thread. Mind you, I don't like it, and I'd be happy if my memory is wrong and the 1994 std did mention that.

But, assuming that we all say that the simulators are to be basically located from each other, then it comes down to the word "location"? Come on! If something is parallel but moves away from the datum (a different distance), its location has changed.

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

http://www.gdtseminars.com

 

[greenimi]

Okay, let's clarify why I have started this conversation again:
In the "back-door location" thread, Evan was looking for some other applications where this "back-door location" could be applicable:

Quote: "I was trying to think of other applications in which the "back-door location" technique would be possible. One of them was controlling a cross hole to be centered on the axis of a shaft. Another was controlling a cylindrical hole to be centered on a spherical feature." Thread posted by axym (Industrial) 7 Feb 12 10:32

So, I have seen a drawing which I thought maybe good candidate and I have created 4 cases based on the original drawing and ask you guys which one (or maybe no one) could qualify for this "back-door technique". Again, as I stated before, academic purpose only. The original drawing I have started from, could have functionality, as Frank mentioned on his post he is looking for,and the subsequent cases created might not have it, that I agree.
Or how the drawing should look like to make it usable for "back-door technique"?
But if everything boils down, now to the datum feature simulators (at least per the latest discussions) how these 4 cases will be affected? (again, use 1994 standard, but as Dean mentioned, and I tent to agree with him, does not make any difference since we are not using translation modifiers anyway)

 

[fsincox]

greenimi,
It is not so much that we don't want to help you learn, it is that I do not want to spread/advocate "bad" practices and until the issue is settled that is what I am in danger of doing.
Your first 3 examples are not the same statements as the one I had. Honestly, at a quick glance, I mistook example 2 for the same condition, you correctly noticed it was different and added case 4. On the face of it case 4 does use the same approach and so based on the final determination of the committee it would be a similar case. Since I brought this issue up, I feel a special responsibility to not mislead anyone and that is why I must express strongly my advocacy for other approaches.
Frank

 

[DeanD3W]

J-P,
You're right that the wording in '94 left a gap a rather wide truck could be driven through... I should have stated that post last night using different words. I think the effective location control an orientation tolerance can provide in some cases should be considered present whether going by the '94 or '09 standard, but you're correct that it's much more explicit in '09.

I don't want to distract from Greenimi's objective here too much, but while the effect we're discussing is definitely "real", I would still always recommend a position tolerance for a more clear spec.

Dean

http://www.d3w-engineering.com

 

[fsincox]

greenimi,
Each of your cases is different and specifies different things. There is a section in the standard they discuss the different framework effects of the specification of a face datum and a diameter datum.. They do this to show that just throwing datums together is not the intent and that each has a specific unique effect. I see many people who just seem to throw datums at features. I suspect you know this but, just in case.
I do not believe you will create the same effect in many other ways, it happens because of the unique relationship of the shaft and slot being “on center” and the way the standard handles the intended “on centerness”. These are it’s own defined rules, ASME’s basic rules of the game so to speak. I haven’t thought about it much, but, it may not even happen under the concept of “Independency” (ISO). I am just starting to explore that concept and think that is why many are opposed to it; it is a basic fundamental game changer.
The ASME invokes a concept of perfection of the datum framework relationship and forces a part to conform to it. It is like an AME, you take an imperfect hole and put a perfect pin in it and measure the pin, this concept of perfection is part of the basic/fundamental foundation of the ASME standard. To ask: “Is this pin straight” is redundancy as it was defined as straight”.
The slot concept is similar in that this world of the perfect slot on perfect center (simulated datums) you create in defining the axis of the non-perfect shaft and rolling the non-perfect slot (datum features) on to center, is established, This perfect world created from the imperfect and allows us to specify a containment that is automatically centered for us and automatically allowed to shrink or grow due to any size variation. It will be forced on center no matter what, due to ASME’s conceptual definition of RFS/RMB (how ASME chose to define the rules, if you will). Positioning is really done by the particular framework specified and not the parallelism, in that sense I guess I am wrong saying it comes from the orientation tolerance. That is also why it doesn’t work for your other examples.
The specific datum definition is also a key here. Many people do not see a secondary or tertiary datum as anything other than the datum itself and thus you get questions like: “how can you define a feature to itself?” While I believe this is mostly true for a primary datum, it is certainly not for a secondary or tertiary datum, now you are back to this image of perfection established from the primary and reflected into perfection as defined again by the rules put in place. Many do not get this, either.
In the end, now, this is also why I can propose the same effect could be achieved with a profile symbol substituted for the parallelism, as long as you are willing to allow profile to be a refinement of toleranced dimensions. The ISO is doing this according to Georg Hensold’s book and I believe the ASME 2009 standard has opened this door now too. .
Does that help any?
Frank

 

[fsincox]

I do not really like the terms "titans" and "fight". I would prefer the terms "people who are interested in understanding better" and "discussion", "back door location" is OK.
Frank

 

[fsincox]

Oh, "confused" is OK too! I am very comfortable with that concept.
Frank

 

[greenimi]

Frank,
You said that each of my cases are different and specifies different things.
That was the intent why I have created them. This is in line what Evan said (in some other threads/posts) the standard should include –more pictures with explicit information
Quote:
“Perhaps the standard should include more "means this" figures. There is a noticeable lack of such figures……..
There could be an example then a series of figures:
This: Means this, and this, and this, and this, and maybe this ...
Evan Janeshewski”

So, if a new concept is being discussed (in our case “back-door location”) a set of pictures to be made and find which one qualifies as a candidate for what we are talking about. Again, I do understand that the datums are chosen arbitrarly (with no direct implied functionality, at least the last three of them), but also I was trying to not “throw datums together”. And every GD and T callout is a valid one (each of them is “legal”), they might be confusing or redundant, but that is the purpose of the discussion, to see which one make sense and which one does not.

I wish I can edit the post to change the title to what you have been requested:
"People who are interested in understanding better the "back door location" concept. Free discussion"

 

[fsincox]

Case 1:
The slot is located from the side and its centerplane refined with a parallelism callout to the thread pitch. The slot is located in this case. Location is done the “old traditional” way. It is process oriented (start from the edge) and one of the things GD&T was meant to replace both literally and philosophically, IMHO.
Case 2:
It relies on a similar framework definition as mine; however, since the part is rectangular in basic shape I doubt you can really tolerate any random orientation of this slot around datum axis “A”. This does allow that. I had a round shaft were orientation is truly irrelevant. So it is intellectually dishonest. In my opinion, to imply orientation of the slot does not matter for this part. Using only datum "A" (the thread) and the slot itself does not control the orientation to the outside profile, unless we specified profile "ALL OVER".
Case 3:
There is no location of the slot at all, implied or stated.
Case 4:
This is the same concept as mine, but again, applied to a different part that is rectangular and will suffer from the same issue as for Case 2.
I am still trying
Frank

 

[fsincox]

pmarc,
Something you may have missed in the original thread was, in the end, It does come down to the actual tolerance numbers specified. Numbers really do matter and most likely the two lower examples you show would NOT pass the stated specification.
Frank

 

[axym]

"Quote:
"Perhaps the standard should include more "means this" figures. There is a noticeable lack of such figures........
There could be an example then a series of figures:
This: Means this, and this, and this, and this, and maybe this ...
Evan Janeshewski""

Greenimi,

I had to go back and look up what thread this quote was from, it's so long ago (2009). You kind of quoted me out of context there, by cutting part of it out. My actual statement was that there is a noticeable lack of such figures for directly toleranced dimensions. Then the part about the series of "means this" figures was a sarcastic comment, that multiple figures would be needed because of the inherent ambiguity of directly toleranced dimensions. I wasn't talking about a series of figures in the same way that you are in this thread.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[dtmbiz]

Datum B being parallel to Datum A appears to be a paradox, considering the FCF
that uses Datum A as the primary datum and Datum B as the secondary datum. By definition
of a DRF they are perpendicular.

The thread has been put to task in a way that seems overly complicated to me.
Cases 1-4 shown , are not fully defined because the varying scenario’s either dimension the
.755/769 FOS dimension to its center plane without a “half” dimension; or appear to rely on a parallel callout to use this “back-door” location.

Both 1994 and 2009 standards state that an orientation callout has a tolerance zone that is “oriented” with the datum referenced in the FCF. It does not say that the oriented tolerance zone is “centered about” a referenced datum.

Therefore there is nothing with reference to the parallel callout that establishes
an allowable tolerance between datum A and the center plane of the .755/.769 FOS.
The sides or center plane (depending on case scenario) are oriented to datum A but do not
positionally locate the sides or center plane relative to datum A.

Another point is that case 2 using parallel FCF with A and B; is orienting a FOS (.755/.769), datum B, to “itself”. Datum A and B being used as primary and secondary datums in a DRF by definition are perpendicular to each other... not parallel.

The position control “establishes the center of a tolerance zone” by use of basic dimensions (or implied 0 or 90). The key is that “basic dimensions or implied basic dimensions that locate” are only invoked with the true position control.

Bottom line is that none of the case scenarios are fully (adequately) dimensioned.

splash...

 

[DeanD3W]

dtmbiz,
1) Please read section 4.5.2 of Y14.5-2009, then
2) draw an axial direction view of the datum feature simulators, then
3) draw a part with a nicely oriented, but poorly located slot, then
4) rotate the part to bring the center of that slot up to fit around the datum feature B simulator...

The sides of the slot are now poorly oriented to the DRF [A,B], so they may not meet the orientation tolerance.

The reason the assertion that Frank brought up is correct is because of the "perfect" relationship that datum feature simulators have to each other. Section 4.5.2 clearly states this.

I'd still apply a Position tolerance, but I think those who are disagreeing with the assertion that an orientation tolerance may in some cases effectively provide a location control (in an indirect manner) are not fully considering how datum feature simulators work.

Dean

http://www.d3w-engineering.com

 

[dtmbiz]

Dean,

The question at hand has nothing to do with datum simulators.
In order to define datum simulators, one needs to define the part.
The part is not fully defined with respect to location of the .755/.769 FOS relative to feature datum A. (with the exception of case 1)

The parallel control is for orientation not location. In the case of a FOS, an orientation needs to be less than the size dimension to be meaningful. A location tolerance does not need to be less than the size tolerance. You are advocating that the location tolerance needs to be .004".... more than likely an unnessecary expensive tolerance. (That is if you buy-in on the idea that the parallelism control also controls location.)

In the cases shown with the parallel control allegedly controlling the location of 2 features is bogus. Delineating 2 features sharing the same center plane simply because they are shown this way pictorially does not address the distance the two features can be offset when produced. This scenario is a classic argument for advocating positional controls with GDT for clear feature relationships vs. part definition without GDT FCF’s.


In the examples shown the two feature center planes can be produced let’s say with an offset of .125" and at the same time
be within .004" parallel to datum A.

I am not sure why one would not simply use a positional control along with the parallelism to remove confusion.

I add the position control to my drawing; you leave the dimensioning as shown on your drawing and we both send the drawings to a couple shops for fabrication. Odds are I will receive parts to print and you will receive phone calls to clarify the print.

BTW,
thats all before you bring the part to inspection and work with datum simulators....

 

[fsincox]

dtmbtz,
Most of the debate is actually a carry over from another thread I posted eariler on a different, but somewhat similar, part:
thread1103-315082

greenmi has asked an additional question on it, here, and is trying to expand on it for his own clarification.

Meanwhile, pmarc has another thread going related to this same question, also.
thread1103-317540

Many of us are thinking of the original condition in these other threads and it is confusing.
Frank

 

[DeanD3W]

dtmbiz,
Are you kidding? Please read my post again.

Did you really read section 4.5.2 before replying?

BTW - They're "Datum feature simulators", not "Datum simulators". If you are able to find any instances of "Datum Simulator" in Y14.5-2009 each is an oversight. In all cases the term should be datum feature simulator.

Dean

http://www.d3w-engineering.com

 

[dtmbiz]

Frank,

Read the threads quickly and it appears that the same point of confusion is being discussed... orientation control "assumed" to also control location...

Dean,

Those threads appear to me that you are in line with my (rather the standards definition) statements that orientation does not control location. I am trying to do this hastily and I missed the word "feature". I agree that the terms are important.
However the ultimate goal is to simulate a datum thru features identified as datums. Datum simulator is not a far reach if you understand the concept; which is that we use simulators to establish datums. A datum is an axis, plane, exact point (1994)...

"2.3.6 Datum, Simulated. A point, axis, or plane
established by processing or inspection equipment,
such as the following simulators: a surface plate, a
gage surface, or a mandrel. See paras. 4.4.1 and
4.4.2."

No I am not kidding.... you dont get it? Really? Really Really???
lil' fun... why not

See the attached for possible results due to lack of proper controls to locate features...

 

[greenimi]

Quote:"
Greenimi,

I had to go back and look up what thread this quote was from, it's so long ago (2009). You kind of quoted me out of context there, by cutting part of it out. My actual statement was that there is a noticeable lack of such figures for directly toleranced dimensions. Then the part about the series of "means this" figures was a sarcastic comment, that multiple figures would be needed because of the inherent ambiguity of directly toleranced dimensions. I wasn't talking about a series of figures in the same way that you are in this thread.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

"



Evan,

I just was trying to support my opinion of having pictures with different callouts to support different interepretations on the same subject is a good idea. And your sugestion and the general idea (to have more pictures) from that particular thread came along with my line of thoughts. I am reading all your pertinent opinions and what I am trying to learn from there, is the concept. And that concept I am trying to apply it consistently to my drawings. Now, I am not very “street smart” to perceive when you guys are sarcastic and when not. I am more concentrated what I can learn from that particular thread and what is the point (which I can use in the future) and not how is the verbiage and how the sentence was said (that is obvious now why I prefer more pictures in the standard and less text). – I don’t want to end up disputing what location means and if it is or it is not the same thing with contained- or if we have to use “absolute location” term instead.

I might have to quoted you out of context there, I agree with you, but I just wanted to make the point that more pictures is a good idea. And on you post I found something to rely on.

Again, I am sorry for this.

I just want to ensure you that you guys provide invaluable information beyond any training sessions can do. It’s a language and it’s a learning process. And I understand even MVP’s not always agree upon a subject and unfortunately the standard does not help (but again I want to get the idea and where are the sticking points, to learn)

 

[fsincox]

We want to be helpful, the problem is this particular issue is really still open for debate and it is very hard to give anyone an "answer" when no answer has been decided.
Frank

 

[DeanD3W]

dtmbiz,
No offense intended, but you seem to be doing everything hastily. I need to say this in the beginning of my post because you don't seem to be reading to end of anything.

You didn't notice that I said in my post that I would apply a Position tolerance anyway, in spite of the clear fact that an orientation tolerance would provide a location control in the specific example described. While what Frank has pointed out is true, I believe, it would be an odd way to go, so it's not something I would ever recommend.

You also didn't notice that the orientation tolerances were applied to each side of slot and they would reference A, the central bore, as the primary datum feature and B, the width of the slot itself, as the secondary datum feature. You left datum feature B off and you applied your parallelism tolerance to the center plane of the slot instead of to each side of the slot.

You have sort of run off the road and you're describing what you found there... Coming back on track and reading about the example that was described by Frank (his post was on January 31st at 9:01) and the sections of the standard that are related the assertion (I think all you need is Y14.5-2009 section 4.5.2) would sure be helpful.

Dean

http://www.d3w-engineering.com