Unequally disposed profile tolerance

[pylfrm]

Another thread (thread1103-401386) has inspired me to ask something I've wondered about a few times in the past.

Consider the following unequally disposed profile tolerance per ASME Y14.5-2009:
[pre]

[box] profile [/box][box] v1 (U) v2 [/box]
[/pre]
And the following three cases:

1. v2 is between v1 and zero, inclusive.
2. v2 is greater than v1.
3. v2 is less than zero.

Case 1 is certainly the main thing I've come across, but I'm curious what you folks think about cases 2 and 3. Are they valid? Are they clear? Have you ever seen them used?


- pylfrm

 

[powerhound]

Case #1 is the only valid case.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech
SSG, U.S. Army
Taji, Iraq OIF II

 

[pylfrm]

powerhound,

Do you feel the standard specifically rules out cases 2 and 3, or that they are just too much of a leap from the usage shown?


All,

The attached image shows two more cases, 4 and 5, based on the alternate practice described in para. 8.3.1.3 and shown in Fig. 8-4. I feel these are perhaps somewhat more likely to be valid than cases 2 and 3. Again, your thoughts would be appreciated.

If this approach is valid, I see a couple potential benefits:
[ul]
[li]Mirrors the commonly used hole/shaft fit system, where tolerances are often either plus/plus or minus/minus from some common nominal size. This would allow basic dimensions of mating features to match, which could help clarify design intent.[/li]

[li]Provides an easy way to specify allowance for future operations that add or remove material uniformly.[/li]
[/ul]


- pylfrm

 

[3DDave]

<rant>

With the already existing example of plus-minus tolerancing, they could have just used the same convention about unequal distribution, but nooooooo. That would be too obvious and the 2009 Committee was all about not being obvious. Their answer is the most bass-ackwards way I can imagine. I know they wanted to keep the first value as the total zone to keep consistent with, well, nothing as the previous technique had no (U) in it. So this was a false consistency.

The math is like this:
4 = +2 <-> -2
4 (U) 0 = 0 <-> -4
4 (U) 4 = 4 <-> 0

In other words, the (U) operator in A (U) B = B <-> -A+B
Without the (U), the operator is A = A/2 <-> -A/2
Perfect.

So if you want the zone to be unequal, only one of those values is displayed in the frame. They wouldn't change plus-minus tolerances to work this way, why do this here?

How about making profile -always- bilateral and deprecating the old method?

Then:
2, -2 = +2 <-> -2
0, -4 = 0 <-> -4
4, 0 = 4 <-> 0

C'mon, too obvious?

</rant>

 

[powerhound]

I answered the question asked. There's no support in the standard for #2 or #3 even by extension of principle.

If the question had been posed using figures such as you provided, then support through extension of principle is there when viewed in the context of +\+ or -\- tolerances. I'm on board with that.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech
SSG, U.S. Army
Taji, Iraq OIF II

 

[axym]

pylfrm,

This is a good question. You're not the first person to wonder about these "unequal unilateral" cases. Overall, I would say that these cases are valid mathematically but I agree with powerhound that there is no support for them in the standard.

A large part of the problem lies in the way that Y14.5 encodes and describes unequally-disposed tolerances. The value after the (U) symbol is described as the "tolerance in the direction that would allow additional material to be added to the true profile". This description only really makes sense if the tolerance zone includes the true profile, and breaks down in cases where the tolerance zone does not include the true profile.

Cases 2 and 4 correspond to "+\+" tolerances for size. Using the (U) notation, Case 4 would be |SPF|0.8|(U)1.2|A|. I would say that this is somewhat confusing, but it works if the 1.2 is interpreted as the amount that the MMB is displaced from the true profile.

Cases 3 and 5 correspond to "-\-" tolerances for size. Case 5 would be |SPF|0.8|-O.4|A|. This is even more confusing, but again it works if the 1.2 is interpreted as the amount that the MMB is displaced from the true profile.

I don't consider +\+ and -\- tolerances to a best practice. But if the standard supports this practice for size tolerances, then it should support it for profile tolerances.

ISO 1101-2012 uses a different encoding method for unequally disposed profile, which I would say is much better. The value following the UZ symbol represents the amount that the tolerance zone is offset, and can be either positive or negative. So the FCF for Case 4 would be |SPF|0.8 UZ [0.8] |A|. The FCF for Case 5 would be |SPF|0.8 UZ [-0.8] |A|. This is so much more direct - there is a zone of a certain thickness, and the zone is offset by a certain amount. I wish that Y14.5 would adopt this encoding format, but I don't think that this is likely to happen.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

Evan, I don't see how your Cases 4 and 5 match up with pylfrm's pictures. Should the numbers after the UZ be 0.4 and -0.4 (according to ISO)?

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

 

[axym]

John-Paul,

No, I believe that the 0.8 and -0.8 are correct.

The value after the UZ is the amount that the zone has been offset from an equally disposed condition. Another way of looking at this is the offset between the middle of the zone and the true profile. In Case 4, the middle of the zone is offset by 0.8 from the true profile and in Case 5 it is offset by -0.8.


Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[Belanger]

cool, thanks

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

 

[pylfrm]

Although it predates the (U) notation, I checked ASME Y14.5.1M-1994. It seems to clearly disallow cases 2, 3, 4, and 5:

A profile tolerance t consists of the sum of two intermediate tolerances t+ and t-.

t+ and t- are always non-negative numbers.

The description used there jives pretty well with the encoding proposed by 3DDave. The ISO encoding makes a lot of sense too. Makes me wonder how the current (U) method came about. Oh well...

Thanks for your input, everyone.

- pylfrm

 

[Belanger]

The current method came about by the input from a leading auto manufacturer with the initials G and M.

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