True Profile Definition

[pmarc]

Hello,

Below is a snapshot taken from the most recent Tec-Ease tip available on their website. My question would be following: Does anyone think that the true profile between points S and T (clockwise) has not been fully defined? I am specifically thinking about close vicinity of points S and T. Thank you!

 

[pmarc]

To orient the hex-shaped hole.

Of course, this could be done without C too (using simultaneous requirement concept - two profiles relative to A and B), but that would be too much for one tip, I guess.

 

[drawoh]

drawoh,

The problem is that a fixture cannot be made ahead of time to verify the location of the radii. If I want to make an optical comparator overlay, I could do it for all the rest of the outline from S to T except the two fillets at S and T. Instead it would depend on the width.

Why do I care about the location of the two fillets? The profile tolerance provides a minimum and maximum outline the radii must be within. Probably, that is all I am interested in. If I care about it, I will call it up.

I see no problems mixing [±][ ]dimensions with profile. Probably, I would apply a [±][ ]dimension to the opposite quadrant of that end radius, and I would replace the "R" with an actual numerical radius with a tolerance. The distance to the quadrant is easily measured. The radius and tolerance controls the form.

I generally do not use the undimensioned[ ]"R". This solves problems on the drafting boards none of us use anymore. The radius called up on 3D[ ]CAD will always update correctly. If I can apply a profile tolerance all around the form, undimensioned[ ]"R" works. Otherwise, I need some form control.

--
JHG

 

[greenimi]

Let me expand a little bit this thread and avoid its “natural death” as I do have a couple of misunderstandings on my own. Again, I use this opportunity to improve my education in this language and hopefully will help somebody else too.

Okay, here are my questions:

Q1: As per my understanding (also confirmed by pmarc) a basic dimension between centers of the arcs (S and T area) is needed in order to make the true profile between points S and T fully defined. The question is why this basic dimension between the centers of the arcs was not needed when the profile was callout as "all around" profile (but again is needed now after the update and when the profile has been changed from “all around” to “in between”)?

Q2: Why a basic width dimension is not shown on either drawings (not the old Tec-Ease revision and not the updated one from Tec-Ease) width of the part (between the hex and the hole-datum feature B)? Why that basic dimension is not needed?

Q3: Can basic dimensions be redundant? In other words, why not adding them “at galore” instead of being insufficient and the true profile being not fully defined? Are they any unintended consequences of such of action, adding multiple basic dimensions?

Q4: Could a reasonable argument be made that even with the basic dimension (between the centers of the arcs) missing, the true profile is still defined (intentionally I did not say fully) in a similar way as notorious fig 8-27 -2009? (And this figure is still maintained in the new draft, so the shown concept might have some validity since year after year, revision after revision is still in the standard, like untouchable one)?

Thank you for your input

 

[Belanger]

Greenimi -- I won't address most of your questions, but there is nothing wrong in the least with Figure 8-27. The standard requires that a true profile be defined with basic dimensions. And that surface is defined by basic dimensions: a basic of zero (it's a flat surface).

There is no requirement that the relationship to the datum(s) must be basic.

That said, a notorious figure that is violating the true profile requirement is Figure 8-18, but I think that's been beat up in other threads.

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

 

[drawoh]

...

Okay, here are my questions:

Q1: ...

Q2: Why a basic width dimension is not shown on either drawings (not the old Tec-Ease revision and not the updated one from Tec-Ease) width of the part (between the hex and the hole-datum feature B)? Why that basic dimension is not needed?

Q3: ...

Q4: Could a reasonable argument be made that even with the basic dimension (between the centers of the arcs) missing, the true profile is still defined (intentionally I did not say fully) in a similar way as notorious fig 8-27 -2009? (And this figure is still maintained in the new draft, so the shown concept might have some validity since year after year, revision after revision is still in the standard, like untouchable one)?

Q2: The radius at the right hand end is defined as a basic dimension. The two sides tangent to this are evidently parallel. If this were my drawing, I would have dimensioned the width, and left the radius as "R". Maybe the radius reflects some sort of design intent.

Q4: I tried drawing this thing out, showing the profiles. It is a bit ambiguous. This is way clearer if there is a dimension explicitly showing the end of the straight 6[±]0.1 section. As drawn, the fillets are tangential to the features are either end. Do you really want to indicate whether or not each and every fillet you show is tangential? Otherwise, your inspector accepts any form that is within the 0.5[ ]profile.

--
JHG

 

[greenimi]

J-P Belanger,
Thank you for your comments. Okay. I understand.
Before your answer, I was wondering if an extension of principle from fig 8-27 has been used on initial/original Tec-Ease drawing to justify/ rationalize usage of ± in combination with profile. Per your answer looks like that's not being the case.

I recently read:

https://www.eng-tips.com/viewthread.cfm?qid=400077

and

https://www.eng-tips.com/viewthread.cfm?qid=320954

and looks like not everyone is in agreement that fig 8-27 is the best thing 2009 has. That's why I brought it up.

And NOOOO...... I do not want to turn this thread about 8-27 again. Leave it alone, please.

All the other questions, are still active!

 

[pmarc]

greenimi,

Let me try:

Q1: In the original version of the tip the 6 width and the R dimension at the left end of the part should have been basic to make the all-around profile meaningful and correct. If that had been done, the basic distance between the discussed arc centers would not have been required as the distance between the arcs in vertical direction would have been controlled by the tangency between the arcs and the sides of the 6 width.

Q2: You should ask Tec-Ease. It is not that the dimension for width C has to be basic. There is nothing wrong in keeping it directly toleranced. But if that is done, the relationship between discussed arcs is no longer basic and that is why the extra basic dimension is needed.

Q3: I think that in general the "beauty" of basic dimensions is that one does not have to worry about overdimensioning too much. Of course, this does not mean that applying too many basic dimensions is recommended practice. I have seen situations where downstream drawing users were seriously confused by that.

If you ask for a situation where addition of basic dimensions could cause unintended consequences, just think about fig. 8-27, mentioned by you, and imagine that the the directly toleranced dimension has been converted to basic. This will change the requirement for the upper face entirely - instead of controlling single line elements only for their form and orientation to the specified datums, the callout will additionally control location. The effect of that will equal to profile of a surface requirement.

Q4: The argument could be made, in my opinion. Imagine that this drawing is converted to Model Based Definition (MBD). In MBD there is no requirement to show basic dimensions (they are taken directly from 3D model), but also there is no requirement that they can't be shown. If basics are shown, someone can still argue that the characteristics that have not been shown should be taken from 3D model. And in fact this is how any CMM inspection software will read the profile requirement between points S and T. It will simply ignore the issue I have raised; it will treat the entire profile between points S and T as fully defined.

This is also one of the reasons why the approach shown in fig. 8-18 mentioned by J-P is vague. In MBD any CMM software that would be able to read the profile callout directly from the model would create 0.02 wide profile tolerance zone that would not only be fixed in orientation and location to the speficied datums, but also fixed in size (i.e. the CMM would treat the diameter 24 as basic). In order to avoid this, the CMM software must be somehow programatically told to ignore the size contraint, and this is where the dynamic profile tolerance zone modifier comes into play (yes, I remember we already talked about its questionable usefulness in case of a cone).

 

[greenimi]

Pmarc and/or everyone,

If that had been done, the basic distance between the discussed arc centers would not have been required as the distance between the arcs in vertical direction would have been controlled by the tangency between the arcs and the sides of the 6 width.

Why your embedded picture it is losing “tangency-ability” when 6mm dimension is not shown basic (but instead ± direct toleranced)?
So, in order to have a true profile fully defined a basic radii is needed. Agreed. And basic radii is shown as 4X R10 basic. Tangency is implied on the right side on the hump (hump over the hex) and not implied on the left side. That’s my understanding of your replay. Not disagreeing, but not understanding why.

 

[pmarc]

greenimi,
Look at the picture below. It has been sanitized so that only basic dimensions that apply to the outline of the part between points S and T have left.

Question: With this set of basic dimensions, would you say that the contour between points S and T has been fully defined?

 

[greenimi]

I would say NO.
What is missing would you ask?
- missing 2X in front of R 10 basic.
- missing R 10 basic for the width of the shank (left RIGHT side of the hump)

Probably (and I think here you are getting at) you say it is missing a basic dimension between S and T points OR between the arc centers of the left RIGHT side of the hump.
Thank you for staying with me. I really appreciate it.

 

[semiond]

I think that on youtube, Tec-Ease have a tip titled: "Do not mix basic and directly toleranced dimensions"

 

[pmarc]

greenimi,
The sanitized definition is lacking a basic vertical distance between the two arcs on the left side of the R10 feature.

'2X' prefix in front of the R10 basic dimension can be added for clarity, but I would not say it is required.

 

[greenimi]

So, you are dimensioning (with basic dimension) the arc location (s) and then the tangency would not be implied by default?
If I leave this arc location basic dimension off, then the tangency would be my default implied?
That is my understanding (but I realize my lack of education in this language) based on 1.8.2.1 -2009 paragraph.

"Where location of the center is unimportant, the drawing must clearly show that the arc location is controlled by other dimensioned features such as tangent surfaces. See Fig. 1-24."

 

[powerhound]

I don't necessarily think there is a problem with the drawing. The points are where the radius nominally comes tangent with the line...wherever that winds up being. I think it's kind of like the R on the left end radius. It's not defined either, but it's supposed to be half the width of the feature, whatever it winds up being. Could this result in a non-tangency? Absolutely.



John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[pmarc]

greenimi, powerhound,

My apologies that instead of giving direct answer to your questions, I will ask my question, but I am not sure how else I could explain the point I am trying to make...

So, if you think that there is nothing wrong with the Tec-Ease tip, should I assume that you also see nothing wrong in following example:

 

[drawoh]

pmarc,

Is that wrong, or just poor quality? If the tolerance were larger than [±]0.25, it would be weird. If it were tighter say [±]0.05, would it not be a refinement of the profile tolerance?

--
JHG

 

[pmarc]

drawoh,
My recent example has nothing to do with tolerance values. The point is that the profile tolerance 0.25 applies all around, yet the true profile of the hole is not entirely defined with basic dimensions. This is the same problem as in the original Tec-Ease drawing (with the exception that in the Tec-Ease example we are talking about profile between points and not all-around).

 

[powerhound]

pmarc,

I see what you're saying now. That was a good example to illustrate your point.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[greenimi]

My apologies that instead of giving direct answer to your questions, I will ask my question, but........

"Tell me and I forget. Teach me and I remember. Involve me and I learn." - Benjamin Franklin

You do not have to apologise. I actually encourage you to do so. And thank you for doing it.

Back to the thread. I do not belive that your latest embeded case (all around profile on the rectangular shape) it is a kosher example. I do not like it nor recommended to be used.
Combining plus-minus dimensions with geometric tolerances is, IMHO, not a good design practice. Looks like you and Tec-Ease agree with the above statement (or at least partially agree) that's why they changed the original case.
Back to my tangency question: when it's applicable and when not? And more important why?

 

[CheckerHater]

My recent example has nothing to do with tolerance values

I am afraid it does.

As the value of profile is smaller than dimension tolerance, profile is acting as a refinement, controlling orientation and form in full accordance with

8.2 PROFILE
...Where used as a refinement of a size tolerance created
by toleranced dimensions, the profile tolerance must be
contained within the size limits...

By the way, where is the image coming from?

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