Straightness of an axis as a refinement of position tolerance 1

[pmarc]

Here is what I have just found in Alex Krulikowski's "Fundamentals of GD&T". Flowchart in figure 4-14 (page 87) shows that specification of straightness control applied to a FOS can be considered as legal when the value of straightness tolerance was a refinement of other geometric tolerance that controlled the straightness of that FOS (e.g. position, total runout, profile of surface, cylindricity).

My doubts are about the presence of position tolerance amongst the exemplary tolerances. Since the straightness applied to a cylindrical FOS controls form of its derived median line (imperfect element) and the position is dealing with the axis of an unrelated actual mating envelope of the FOS (perfectly straight element), I see no connection between straightnees and the position of the FOS.

Example: imagine a pin that is controlled by positional tolerance relative to typical DRF created from 3 orthogonal surfaces A, B, C, and additionally (for some reasons) its 'axis' is controlled by straightess tolerance. I see no reasons why the straighthness tolerance value couldn't be greater than the value of positional tolerance. A banana-shaped pin with very large 'axis' straightness deviation could be well inside smaller positional tolerance, at least in my opinion.

Is Alex's statement correct? Any thoughts?

 

[Madhu454]

Hi PMARC,

I completely agree with you, the positional tolerance does not control the straightness of the surface elements of the hole, the same error can refect to the median line created from the hole surface. where as the axis created by AME can still lie within the specified position tolerance zone.

But if the drawing is created as per ISO-1101, then straightness value should be refinement of positional tolerance value. Because in ISO not the axis should be within the tolerance zone, it is the median line should be within the tolerance zone.

I also had same question, lets see how others comment!




Madhusudhan Veerappa
Mechanical Engineer

 

[pmarc]

Madhusudhan, well noticed about positional tolerance interpretation in ISO 1101. Actually IMO, this is one of the most important differences in interpretation of ISO and ASME GD&T standards.

 

[fsincox]

It seems odd to me that while trying to kill concentricity, ASME has now made ir the only way to control the coaxial location of the actual feature axis instead of an AME axis or using some combined control (runout). Seem like I might want to use it more now not less. Again, I find myself glad the ISO does not by it all "hook, line & sinker".
Frank

 

[DeanD3W]

Frank,
What Y14.5 needs to include, in addition to Position, is what some have proposed be named "Local Position", which would apply to a feature's median line or plane... I hope it will be added in the next revision or two.

I would prefer that Concentricity and Symmetry be deleted. Local Position would apply to a wider variety of features (tubes with bends, for instance) and capture functional requirements better, in my opinion.

Dean

http://www.d3w-engineering.com

 

[Madhu454]

Hi Pmarc,

Not only the Alex Krulikowski's "Fundamentals of GD&T" text book says this. While going through the ASME 1994 std I found this in Page 202. It says this way

" A straightness tolerance on a FOS, normally permitting a violation of the MMC boundary, is not allowed when used in conjuction with an orientation or position tolerance. in such a case, the specified straightness tolerenace value shall not be greater than the specified orientation or position tolerance values.

I understood the above statement this way
1) Genarally by applying staightness to FOS, the Rule#1 can be violated.
2) If staightness to FOS used in conjunction with say Position, then the Rule #1 (perfect form at MMC) cannot be violated.
3) Also the staightness tol value should be a refinement of Position/Orientaion.

But really I did not understand Y it cannot violate the Rule#1?

Evan I did not understand how the median line and axis of the FOS related to each other (as you explianed "A banana-shaped pin with very large 'axis' straightness deviation could be well inside smaller positional tolerance")

The one reason may be - the WCB created by Staightness of FOS should not be graeter than the the virtual/WCB of the FOS due to position/orientation tol. What do you think?

 

[axym]

Madhu,

I believe that the confusion over how Straightness interacts with Position and orientation tolerances is primarily caused by the two "interpretations" that Y14.5 defines (surface interpretation and axis interpretation). Each interpretation results in a different type of interaction.

If the Straightness tolerance is referenced RFS, it controls the derived median line (DML) of the feature. If the Position tolerance is referenced RFS, it controls the axis of the surface. I agree that a feature's axis and derived median line are not directly related to each other. So as pmarc says, we should be able to precisely control the axis of a banana-shaped pin. In this context, the rule requiring that the Straightness value not be greater than the Position tolerance does not make sense.

Things change if the Straightness tolerance and Position tolerance are referenced at MMC. Each tolerance would define a virtual condition boundary that the surface cannot violate. In this context, the Straightness tolerance needs to be a refinement of (i.e. be smaller than) the Position tolerance. I believe that the reasoning for the rule in the standard is based on this MMC case.

So the requirement for the Straightness tolerance to be smaller than the Position tolerance does not make sense for tolerances at RFS, where the "axis interpretation" applies. But it does make sense for tolerances at MMC, where the "surface interpretation" applies. Unfortunately, Y14.5 does not do a good job in describing when and how the different interpretations apply, especially for Straightness tolerances.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Madhu,

Evan (axym) nailed it.
Y14.5-2009 is more specific on that (p. 5.4.1.2), however still using similar statements.

 

[Madhu454]

Thanks Evan, for the clear explaination.