"unit basis" for profile of a line 2

[concerta]

Hello All,

I am very beginner at GD&T. Have an urgent question, please help me guys

The question is: Is unit basis applicable for profile of a line?

 

[GDTcoach]

Easy now, CheckerHater. From the published Standard at the end of 8.3.2.2: NOTE: A profile per unit length, similar to that shown in Fig. 5-4 for the control of straightness, may be used to control abrupt transitions that occur when profile tolerances are specified on adjoining segments of a feature.
Not clear how you can say that profile on a unit basis is disallowed. It is not "wishful thinking" as you stated. The section reference is a typo that I will correct immediately (if you never do anything, you never do anything wrong). Yes, it talks about using the concept to control abrupt transitions but it does not disallow using it elsewhere. In fact, I have a customer who makes knee implants who has used it very effectively. Their scanning software is capable of inspecting the profile of a surface on a unit basis tolerance. Don't want a snappy knee joint.

 

[CheckerHater]

GDTcoach,
I am glad to see some clarification.
The note still part of Para. 8.3.2
It only applies to “abrupt transitions” that occur within non-uniform tolerance zone (You quoted it in your own post), not to abrupt changes in profile itself.
It does not universally apply to the entire Paragraph 8.
But let’s go back to your tip.
When I “imagine a flexible coin that could conform to the basic contour to visualize the tolerance zone” I see that coin sweeping exactly the same area as good old vanilla profile tolerance zone.
This is why I see idea of per-unit requirement meaningless when applied to profile.
Per-unit requirement works well with datumless controls where feature is related to itself, where it probably belongs anyway.

 

[pmarc]

CH,
If it matters to you, I fully agree with Don on this.

Your argumentation that the note from 8.3.2.2 applies only to 8.3.2.2 is not convincing. Read para. 8.3.1. It says that: "Since the surface may lie anywhere within profile boundary, the actual part contour could have abrupt surface variations. If this is undesirable, the drawing must indicate the design requirements, such as rate of change and/or blend requirements." Don's tip is exactly the indication of required rate of change.

I also do not agree with the statement: "Per-unit requirement works well with datumless controls where feature is related to itself...". Case #3 from my sketch is exactly showing per-unit requirement related to datum plane A - notice that each smaller tolerance zone is parallel to A. This is what isn't required by profile without datum references, thus my remark: "1. Case #2 and Case #3 - smaller tolerance zones are shown parallel to datum plane A, while in reality it does not have [to] happen."

 

[GDTcoach]

Well, CheckerHater, as you said, and I agree, "Per-unit requirement works well with datumless controls" which is exactly what I show in the Tip. Since the unit basis control does not have a datum feature reference, it is free to float (in location and orientation) within the larger profile tolerance that is controlled relative to a datum. Your comment that "it only applies to abrupt transitions" is like saying that tangent plane only applied to orientation tolerances because that is where it was shown in the 1994 revision. That sort of reasoning made us add the note to the end of the current tangent plane section. For the readers of this thread, often it is necessary to use extensions of principle to clearly define design intent. You will not always find your exact application in the Standard. Ask yourself if what you have said on a drawing will have one clear meaning to anyone knowledgeable of the Y14.5 standard. Often people ask "Where does it say I can do that?" A better question might be "Where does it say I cannot do that?"
As a member of the Y14.5 committee and Section Sponsor of Profile, I am reluctant to participate in these discussions because some folks just like to take shots at the committee and our efforts.

 

[GDTcoach]

Hi pmarc, we were posting at the same time. You bring up a good point regarding having a datum feature reference to control orientation.

 

[CheckerHater]

Pmarc,

If Note was meant to apply universally to all profile cases, it would not be added to Para. 8.3.2, but to Para. 8.3.1 right after it said “the drawing must indicate the design requirements”.

I m not happy with the way Y14.5 handled it. I said it and I can repeat that it looks like afterthought, last minute adjustment, but face it:
When explaining how to deal with abrupt variations in profile standard DOESN’T SAY “use per-unit principle”. It says what Para.8.3.1 says.

I am still not convinced.

We had this discussion before. If it’s not in the standard, it has to be explained on the face of the drawing.

 

[pmarc]

And I am not going to further convince you - I do not want to end up searching forum history for how many times we both agreed that "extension of principles" is acceptable or not depending on circumstances. That is too tiring for me.