Spinoff2: Using orientation to control position 1

[fsincox]

The whole idea of using orientation to control location has also come up in my past work as the “GD&T guy”. A very well respected senior inspector proposed the use of parallelism to control the location of shaft keyways. His idea was that you would line up the shaft diameter (the primary datum) and then roll the keyway, itself (the secondary datum) to center and proceed to measure the resulting parallelism of the opposed sides of the keyway. His rational was that the resulting parallelism of the sides of the keyway in this representation of the installed state was important to the key not working its way out under heavy loading. I have to admit I did not feel it was technically illegal, what do you think?

 

[CheckerHater]

Frank,

You derive datum from sides of the keyhole and then you control sides of the keyhole to datum .
It looks too much like self-reference to me. That's why on my picture I invoke datum [A-B] which is neither [A] or , so it's a legal loophole of a sort.
Still, I am not sure about advantages; how is it better than specifying Location, and how would you gauge the part anyway?

 

[fsincox]

CH,
I was not the originator of the idea, as I mentioned. I did have to calculate the alowable position tolerance implied. I believe the gentleman I worked with felt it was easy to set up and inspect, otherwise I do not think he would have proposed it. He was a senior level inspector on our company GD&T committee. I believe he also served on an ASME committee for our company relating to inspection. I was the chair and engineering representitive and had a more "is this legal" kind of role.
I do not think it is any more of a self reference than a shaft with (2) bearing diameters using A-B as a datum framework. I think you are not use to it, I really don't see it as illegal though. I was very interested at the time in pushing orientation to a frameworks not just: "the perp to A", "perp to B", stuff.
Frank

 

[drawoh]

CheckerHater,

I am looking at your drawing.

Why can the slot not just be parallel to datum[ ]A? The cylinder defines the vertical. That is all your parallel specification is controlling?

JHG

 

[ewh]

That was my thought also; wouldn't you have to hold the parallel tolerance twice as tight if used on both sides, or is this being accounted for in the slot width tolerance?
Hope that isn't too far in the weeds...

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[CheckerHater]

drawoh,

The way I see it, [A] by itself only defines the axis (and infinite number of planes crossing it).
To pick one plane ("vertical") you need second, "clocking" feature, hence use of [A-B].
Keyway may be off-center (like in Fig 4-32), but still parallel to axis (datum [A]), that's why I believe [A] alone is not enough.

 

[fsincox]

I actually wanted position, as it was simpler for me to conceptualize and calculate and more "by the book". He was a senior respected person and I was basically handed this and asked: ”can we do this?”. I did buy into it after a lot of thought and discussion, though, as you can see, here.
Frank

 

[pmarc]

Frank,
I agree with you that assigning parallelism callouts to both surfaces of the keyway relative to a centerplane derived from keyway's width does not fall under datum self-referencing, but I can't agree with your statement that in this case parallelism controls location through the back door. Look at the attached picture - I hope it explains what I mean.

http://files.engineering.com/getfil...8885-ab1b90e098fc&file=parallelism_keyway.JPG

The point is, parallelism never locates - not through the back door, not through the front door, never.

 

[fsincox]

pmarc,
Your picture and callouts are perfect. By calling out B as the secondary datum don't we forced it to be rolled/aligned to the center to establish the proper inspection setup?: Thus, forcing the other parts you have kindly provided to skew their keyway sides. Isn't it actually parallel in an A (primary), B (secondary) complete framework?
Frank

 

[drawoh]

fsincox,

Here is my take on it.

Datum[ ]A provides location and a parallelism reference. Datum[ ]B provides clocking for any subsequent features. It is of no relevance to this drawing.

The positional tolerance actually controls parallelism, so the separate parallel tolerance is required only if it is more accurate. There is no point doing this unless the diameter is accurate. The slot probably is not a good reference unless it is more accurate than I have shown.

As noted by pmarc, the parallelism specification has no effect on location.

JHG

 

[CheckerHater]

Frank,

You can say there is implied "zero" dimension between cener of the slot and center of the shaft.
Unfortulately it is NOT "basic zero", but rather "titleblock zero", so nothing is "forced".
And how many decimal digits implied zero has, anyway?

 

[pmarc]

Frank,
Answering to your question: "By calling out B as the secondary datum don't we forced it to be rolled/aligned to the center to establish the proper inspection setup?", the answer is no. There would have to be a positional tolerance of keyway's width relative to datum axis A to keep the keyway centered at the shaft's axis.

As for drawoh's sketch - the idea is fine, but notice that by controlling parallelism of slot's centerplane you are not limiting mutual parallelism of its side faces. Parallelism callout says that the centerplane of actual mating envelope of the slot has to be within two parallel planes 0.05 apart, however the faces can be symmetrically divergent along the length of the shaft, so the actual mating envelope will be even perfectly parallel to datum axis A, but the only thing that will control mutual parallelism of the faces will be size tolerance of the width.

Uff... I hope I did not make the thing even worse.

 

[fsincox]

I thought if something is shown on center (and no additional dimensions given)it is understood to be implied to be on center? I agree that there is no tolerance that says how close it has to be and no decimal places to zero.
Many of the drawings in the standard that position things on center rely on an impled basic zero, coplanarity also relys on a impled basic zero.
Frank

 

[pmarc]

Frank,
That is true. The difference is that for position callouts basic zero distance dimension applies, and for orientation callouts basic angular dimensions applies. So in case of parallelism it will be basic zero degrees, not basic zero millimeters or inches.

 

[DeanD3W]

Go pmarc Go

Frank - Your senior inspector friend was so far off in the weeds on this one that he ended up in the swamp. There is absolutely no location control to be found where he was looking. He had/has a misconception regarding orientation controls.

Dean

http://www.d3w-engineering.com

 

[fsincox]

drawoh,
I am sorry I do not agree, so far, You admit that you know what to do with the A (primary), B (secondary) framework for other features, It is in fact the same thing I or others would do. The establishment of a shaft diameter and key for clocking is in the standard as a common framework for these type of parts. I do not see why you know how to "clock" to the key "after". I was taught that process should have no relevance to how you establish a datum framework; I believe you should establish the framework the same all of the time. That is basic to the concept of repeatable measurement. The drawing is a description of the final product not a process of making it. I feel the wording in standard supports the concept that process is irrelevant.
pmarc,
It is actually only parallelism that is being checked so I am not following the problem with the parallelism. Part of the parallelism error may be the result of what we would call location error because we may happen to notice the sides of the key are actually parallel to each other.
I will admit, I do have a stake in this because you could say I vetted it, for right or wrong. So I am not giving up that easy.
Frank

 

[fsincox]

CH, pmarc & drawoh
I do want to thank you for your work on the illustrations. I haven’t got the posting/hosting part figured out yet. This last illustration by pmarc is correct for better or worse.

 

[axym]

Frank,

I agree with pmarc and Dean. A Parallelism callout isn't going to give location control, no matter how creatively it's applied.

Here's another way to look at it, that brings in some ideas on the tolerance zone mechanics of location. A Parallelism tolerance zone is different from a Position tolerance zone in one important respect - the Parallelism zone is allowed to translate relative to the datum reference frame, and the Position zone is fixed. This is what gives the Position characteristic (and the Profile characteristic) the extra power to control location, that the Parallelism zone does not have.

So you can establish a datum reference frame using whatever datum features you want, even the considered feature itself. But after the DRF has been established, the Parallelism zone can still translate in any direction relative to it. This applies whether basic dimensions are shown or not. The zone cannot rotate relative to the DRF (this is what gives orientation control), but it can freely translate. This translation can give the results shown in pmarc's illustration.

I believe that this agrees with the descriptions the others have given.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Yep, nicely stated, Evan.

 

[fsincox]

I think you guys do agree with each other, too.
Once an orientation zone is established by a point on a surface are not ALL other surface points to be LOCATED in a zone of width equal to the stated Tolerance? This is the concept that forces flatness control from orientation also. That provides the required control here. It is still only an orientation control, the sides are allowed to float within the greater size tolerance zone. Once a point on each side is established there are (2) zones of fixed width, parallel to the axis/framework that all of the surface point must fall (be located within). Is there not?
Frank

 

[fsincox]

The standard ASME Y14.5M-1994, states:
6.3.1 Form and Orientation Tolerance Zone.
“A form or orientation tolerance zone specifies a zone within which the considered feature, its line elements, its axis, or centerplane must be CONTAINED.” The word “contained” to me means located within, I am certain the main reason they did not use a term like location is exactly the problem we are having here, people would be confused and say: ”orientation locates”? You must ask yourselves what the word contained means to you, to me it forces location!