Perpendicularity 1

[metaldork]

Can anyone please clarify if a perpendicularity call out is controlled by a basic dimension or not? In other words is the tolerance zone centered on a basic dimension?

 

[dtmbiz]

Not sure how you are supporting your "point" assertion... just dont get your point

What is "is" ?

Good luck CH...

 

[dtmbiz]

CH,

Since I do agree with much that you write, I still have hope for you....

1.3.22 Dimension
dimension: a numerical value(s) or mathematical
expression in appropriate units of measure used to
define the form, size, orientation or location, of a part
or feature.


1.3.27 Feature
feature: a physical portion of a part such as a surface,
pin, hole, or slot or its representation on drawings, models,
or digital data files.

No points though.... I will concede that points do get dimensions at times...

however not in normal dimensioning of a rectangular block...

 

[Belanger]

A feature of size must have directly opposed points. So yes, pretty much any inside/outside idea can be a FOS, if the surfaces are directly opposed.

But also, for something to be a FOS, it must associated with a directly toleranced dimension. So if we use Fig. 4-9 in the standard as a random example, the slot in the left side of that part is not a FOS.

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

http://www.gdtseminars.com

 

[pmarc]

Are we still talking about Perpendicularity?

 

[CheckerHater]

We are talking about your sketch.

 

[axym]

pmarc,

I agree with all of your statements about the Perpendicularity callouts.

CH,

Ambiguity may not be a swear word, but in Quality I would say that it's at least a dirty word. If different interpretations of a spec are possible, then different assessments of conformance to the spec are also possible. We all know how nasty this can get.

dtmbiz,

If the as-designed relationship is 90 degrees, then Perpendicularity is the appropriate orientation tolerance to specify. The Perpendicularity tolerance zone is always at exactly 90 degrees to the datum, that never changes.

When pmarc said that the Perpendicularity tolerance could be infinite, this was just to emphasize that the Perpendicularity tolerance is completely independent of the size tolerance in this case. I don't think he's suggesting that an arbitrarily large Perpendicularity tolerance should be specified. That said, there are cases in which the as-produced tilt angle of the feature can get very large, perhaps even close to parallel, and still conform to the Perpendicularity zone. This can occur when Perpendicularity is specified on features in very thin material.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[CheckerHater]

If only things were that easy.
Take a look at the picture:

http://www.aceseal.com/images/tech_5_radical.gif

"As designed" angle is 90 deg., which doesn't help.
What exactly is "F" measurement?
Is the groove "feature of size"?

 

[axym]

CH,

That drawing is a good demonstration of the disadvantages of dimensional tolerancing compared to geometric tolerancing. Dimensions are easy to define on perfect CAD geometry, but are ambiguous on imperfect real geometry.

What exactly is the "F" measurement? Good question. It's the distance between two surfaces that are not necessarily flat and not necessarily parallel. The F dimension could be defined in more than one way on an as-produced part.

Is the groove a "feature of size" ? If that drawing is all we have to work with I believe the answer according to Y14.5 would be no, because the F dimension is not directly toleranced.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[dtmbiz]

Axym,

I think you are missing it. Depending on the size of the tolerance zone, the angle from the bottom of the part to the top of the part across the tolerance zone will result in an angle. Once the tolerance zone is large enough to allow an angle greater than 1 degree lets say, and then maybe up to 5 degees for example; that surface would no longer be a right angle to datum A.
Of course the tolerance zone is perpendicular to Datum zone within the limits of datum simulation. However if the tolerance zone is large enough you will not have perpendicularity.
Forget infinity, lets look at a .25" tolerance zone yielding over 7 degrees of angle. That result is not perpendicular to me.
Basic geometry.

 

[dtmbiz]

CH,

Don't see how you compare the original sketch and an o-ring groove.

Your example shows a 90deg angle with a max 5deg.

The sketch pmarc offered uses a perpendicular control which suggests to me that something very close to a right angle is required for design function.

So we can now call perpendicularity achieved on a thin part that has almost a parallel relationship?

 

[pmarc]

dtmbiz,
First of all let me say that Evan exactly described my intentions of posting the sketch - it was only a theoretical excercise, without going into considerations about functional requirements and all that stuff. And in the light of that, "infinity" is absolutely correct answer, though of course no one will ever specify infinity on a print. But like I said - it was only a theoretical question.

Now, to your last post. No offence, but I am afraid the sketch you posted and all your assertions stem from the simple fact that you are mixing "perpendicularity" with "perpendicularity error". You already cited what "perpendicularity" is. I will just emphasize it is a perfect condition in opposite to "perpendicularity error" which tells how far from this perfect condition the actual surface is. Of course in order to have the as-produced surface within the spec., this error has to be less than or equal to perpendicularity tolerance value specified in perpendicularity FCF. But the point is the actual surface of the part can be at any angle to datum, even very close to 0 degrees and still be inspected for conformity with perpendicularity requirement, and not angularity or parallelism.

 

[axym]

dtmbiz,

I was thinking along the same lines as pmarc and was about to write a response, but he beat me to it.

Y14.5 defines several characteristics in terms of the perfect condition, which many people (including me) find a bit counter-intuitive. Perpendicularity is the condition in which the surface or axis is at a right angle to the datum plane or datum axis. Perpendicularity is only achieved on idealized entities in the drawing or model. No real feature is ever Perpendicular, by the Y14.5 definition. In reality, As pmarc mentioned, all real parts have "Perpendicularity error" and so there must be a Perpendicularity tolerance.

But the Perpendicularity tolerance does not directly control the angle. It will indirectly control the angle at which a perfectly flat as-produced feature could be tilted, but this control will depend on the ratio of the height of the feature to the width of the zone. The taller the feature, the less the feature could tilt and still conform to a given Perpendicularity tolerance. For very thin parts and thus very short features, a seemingly small Perpendicularity tolerance may allow a significant (even extreme) angular tilt. This is especially true for Perpendicularity tolerances on cylindrical holes referenced at MMC.

So yes, the way Y14.5 defines things, the as-produced part surface could be almost parallel to the datum and still conform to the Perpendicularity tolerance.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[CheckerHater]

Thank you guys,
I apologize for confusing posts; as I was trying to communicate with 3 people at once, it was hard to present my point in concise manner.
pmarc's sketch just triggered some old controversies and never fully satisfied doubts.
It is popular opinion among GD&T crowd, that regular dimensions are no good for anything without geometrical control. And yet it looks like dimensions are here to stay.
Different standards have different way to deal with it. ISO adds missing requirements by means of 2768 (ISO has its own drawbacks, this is not the place for another discussion); ASME has Envelope req't (Rule 1) that indirectly controls some geometrical requirements for features of size.
The problem is: not every feature is a feature of size, and we are not always sure it's the FOS we are dealing with.
Is diameter symbol indication of FOS?
Is chamfer FOS?
What if we dimension chamfer with diameter?
When we apply dimensions to the drawing, do they always mean what we think they mean?
I had my first discussion with manufacturing guy about checking O-ring groove back in '85 or '86. The picture in the standard did not become any more clear ever since. BTW, the dimensions are toleranced thru general notes, so if that's the only obstacle to calling groove feature of size, no worries.
So the sidewalls of the groove are square by design (but may be not) and dimension "F" is probably measured between edges, which, in turn, are rounded, so we have to resort to "virtual sharps".
Now, Is pair of imaginary features FOS?
This could be continued but I think I better break the chain of thoughts started by pmarc's sketch and keep quiet for a while.
Sorry for hijacking the thread.