Does Dia Tolerance Control Straightness 1

[swertel]

I have a tube with an inner diameter dimension consisting of a +/-.005 tolerance.
Does that tolerance also control the straightness of the tube?

From my interpretation, if the ID was a datum, then the theoretical datum surface would be a cylinder that fits within the tube. That implies straightness must be maintained. But, what I don't remember is if the same applies to the physical surfaces. I'm afraid that I will end up with a bent tube whose diameters at random places fall within the size tolerance, but will not functionally work with its telescoping mating tube.

(I'd love to change the print to include a straightness callout just to be sure, but this is a build-to-print job. I don't control the design.)

--Scott
www.aerornd.com

 

[swertel]

I also forgot to mention that bent tube, due to the forming process, will not have a consistent external and internal radii. Therefore, the diameter at the tube is likely to not be as defined. How much the bends get egg-shaped is up to the forming (bending) method and the skill of the operator.

I bring this up for 2 reasons:
[ol 1]
[li]Some quality groups are anal retentive and will reject the tube because it doesn't meet diameter tolerance at the bends. Be ready for that argument.[/li]
[li]Because I used an exhaust pipe as my example, the constriction around bends is something you want to consider if you are modeling CFD. If you are running CFD with the perfect CAD file as the input geometry, it won't give realistic results.[/li]
[/ol]

--Scott
www.aerornd.com

 

[chez311]

djhurayt,

General consensus has nothing to do with it. It simply comes down to whether or not rule#1 applies. Y14.5-2009 sets out a clear set of situations for when this happens - again I would highly suggest reviewing the aforementioned applicable sections of Y14.5. The below are the two cases:

1) Rule #1 applies => perfect form (straightness) at MMC size required
2) Rule #1 OVERRIDDEN => perfect form (straightness) at MMC size NOT REQUIRED

Paragraph 2.7.1 , 2.7.2 , and 2.7.3 all lay out clearly the situations where rule #1 applies or not.

 

[chez311]

djhurayt/swertel,

Rule #1 applies to each INDIVIDUAL REGULAR feature of size on a part - the situation that swertel laid out (a bent tube which is specified as such - ie: segments of straight/curved sections with dimensions applied to each) is NOT a regular FOS so rule #1 does not apply.

In your example rule #1 could apply to each individual applicable segment which is a regular FOS, but not the part as a whole.

 

[Belanger]

djhurayt and chez311 -- also don't forget paragraph 1.4: "Unless otherwise specified, all tolerances apply for full depth, length, and width of the feature."
One could argue that the "feature" ends when the tube changes from a straight length to a bend.

 

[djhurayt]

chez311, that is the piece that I was not connecting "segments of straight/curved sections with dimensions applied to each) is NOT a regular FOS so rule #1 does not apply"

I see the difference now.

 

[chez311]

JP,

Is it even really necessary to reference paragraph 1.4? Without explicitly stating a relationship between individual features per 2.7.3 or 2.7.4 I would think the distinction between individual regular FOS or between a regular FOS and any other individual feature/irregular FOS would be enough to tell where a feature ends. Rule #1 clearly states it only applies to individual regular FOS.

 

[Belanger]

chez311 -- well, I looked at the paragraphs you referenced (2.7.1 , 2.7.2 , and 2.7.3), and they all refer to Rule #1, which speaks about "the limits of size of a feature" (and yes, FOS).
I wanted to focus on what is meant by "feature." Thus my comment.

 

[pylfrm]

Paragraph 2.7.1 , 2.7.2 , and 2.7.3 all lay out clearly the situations where rule #1 applies or not.

I'm not aware of a clear procedure to determine whether a part is "subject to free-state variation in the unrestrained condition". This seems like it might be relevant to OP's situation.


pylfrm

 

[3DDave]

All parts are subject to free-state variation, if one measures closely enough. The default unstated interpretation seems to be that if the part deflects under some amount of gravitational acceleration more than the user might like, then that qualifies.

What about the second line of 2.7.2a?

"Unless geometric tolerances are specified on the drawing of a part made from these items, standards for these items govern the surfaces that remain in the as-furnished condition on the finished part."

Seems like one can apply geometric tolerances on even as-furnished surfaces, as only in the case that no geometric tolerances are applied do the external standards govern the as-furnished surfaces.

 

[chez311]

JP,

I see - I was focusing more on the fact that rule #1 applies to individual regular FOS, and when reading more closely I realized that perhaps it could be mentioned that multiple regular FOS could be simultaneously required to have perfect form at MMC (rule #1) by a few methods laid out between 2.7.4/2.7.5 (my apologies - I initially mentioned the wrong sections) including explicitly stated coaxial/symmetric features, a custom note, and <CF> continuous feature symbol/note. Hence why I said that unless explicitly stating the relationship between multiple regular FOS, rule#1 only applies to individual regular FOS.

If you're mentioning 1.4 and probing into definition of a "feature" can I assume that you believe that 1.3.32.1 does not clearly enough define a regular FOS? Lets take the example put forth by swertel with the exhaust pipe. If there is a diameter applied to a straight cylindrical section and this meets a bent/radiused section is there any interpretation that one could take that to be a single (individual) regular FOS?

(edited for clarity)

 

[Belanger]

Hi chez311 -- some might interpret that as you mentioned. But if Rule #1 only applies by default to a regular FOS, then according to a strict reading of 1.3.32.1, as soon as a straight pipe gets to a bend it is no longer part of the same regular FOS.

I might actually take it back to the general word "feature," because I known that very topic has been somewhat debated on this forum in years past. The classic example is when a profile tolerance is applied to a strange contoured shape (think of a plate or block with a curvy edge). For clarity, we would indicate the between symbol "between X and Y." But if that were not given beneath the FCF, there would be 3 interpretations:

- the profile tol only applies to that radius; as soon as it transitions it is technically a different feature
- the profile tol continues until a sharp corner is reached
- the profile tol applies to the curve as long as basic dims continue to define the nominal curve

I might put myself in the first group. But I know that others disagree.

 

[chez311]

But if Rule #1 only applies by default to a regular FOS, then according to a strict reading of 1.3.32.1, as soon as a straight pipe gets to a bend it is no longer part of the same regular FOS.

This is the interpretation I would go with. Unless someone can present a compelling argument otherwise I think that deviation from this creates more questions than answers, don't you think?

The classic example is when a profile tolerance is applied to a strange contoured shape (think of a plate or block with a curvy edge). For clarity, we would indicate the between symbol "between X and Y." But if that were not given beneath the FCF, there would be 3 interpretations:

I know you're trying to make a point about the definition of the word feature, but its a little hard for me to make the connection to regular FOS. Whenever we start talking about irregular FOS or a feature that falls into neither of those categories I think things become inherently more fuzzy and less clear.

As far as the 3 interpretations, I've previously given that some consideration and I'll be honest I'm not 100% in a single camp. I think I would put myself in the second group where the profile ends with a sharp edge, which I know leaves some room for whether corners/edges which have a very small radius/edge break constitutes sharp. I think if I were the inspector I would try to go with whatever is most conservative during inspection but I do know that whenever I specify on a drawing I try to eliminate any ambiguity by using "between" notation if the all around/all over symbol is not applied.

As I said though I think the analog to a regular FOS is a tenuous one. Whereas the interpretation of how your profile tolerance applies changes only which part of a profile is held to the specified profile tolerance, interpretation of the term feature such that one deviates from a strict reading of 1.3.32.1 I think introduces a lot of speculation and stretching of the rules. If a feature with both straight and curved sections is to be considered a continuous regular FOS, how does that change how surface or DML straightness/flatness are interpreted?

 

[chez311]

pylfrm/3DDave,

I agree pylfrm that as you noted there is not a clear procedure for determining whether a part is subject to free state variation - I guess I should have omitted the word clear, the standard does at least lay out the situations where rule#1 applies or not. Perhaps whether the exact definitions and application of those situations is less clear**.

That being said, I would probably agree this is something which the part and tool designer should take into account. It was already mentioned the deflection of the tool/mandrel over which this tube is fabricated needs to be considered. Since this is also a long thin walled tube (albeit made from a composite - likely something very light and stiff like carbon fiber) flexibility of the finished part may also be of concern - this may be something that swertel and his QC team may want to discuss with the design responsible party to determine what is acceptable and whether restraint during measurement is necessary and what that would look like.

Of course the fact that perfect form at MMC may no longer apply in the free state would have to be taken into consideration.

**At least related to free state variation I think we can all agree that some level of ambiguity here is purposeful as it is up to the designer not the ASME committee to determine if the combination of tolerances, material (rigidity/modulus), assembly method/forces, and application itself (ie: the forces it will see in service) dictate if deflection needs to be considered and how "restraint" should be defined to simulate this deflection. As 3DDave noted there is no such thing as a perfectly rigid part even though the standard differentiates between "rigid" and "non-rigid".