Relax tolerance on all around profile 1

[greenimi]

A pentagon shape is controlled by .005 al around profile (no datum). The mating part is a similar shape part. The product engineer is insisting to add another callout to lessen the tolerance on two sides of the pentagon. In other words, from 5 sides, three of them to be in .005 and the other two in .010 AND still use all around profile. How can I legally, do that?


I realize that I’m speaking somewhat theoretically, but I would like to know what options/tools Y14.5 can offers.


I was thinking to use in between profile (name the corners A, B, C, D and E and use profile between A and B, and another
profile between B and C, etc), but in this case can I use all around profile (for assemble-ability with the mating part)?


Thank you

 

[axym]

Hi All,

The discussion about the meaning of "locate" illustrates some of the potential problems with this kind of spec. There are no rules for the application of directly toleranced dimensions to anything other than features of size. So we are left to figure them out for ourselves, which can result in different opinions. The following question could be asked:

What does the center point represent on a real part? Is it the center of an arc derived from the as-produced surface, or is it part of the definition of the Profile tolerance zone? I believe that this could be argued either way.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Evan,
Regardless of whether you agree with the commments I made about figure 2 from the newsletter, I thank you for your last reply.
Why to use a dimensioning technique for which there are no rules in the standard? Especially that there are methods to accomplish similar effect without any ambiguity.

I wish future editions of Y14.5 put slightly more emphasis on showing why certain dimensioning and tolerancing methods are bad from geometric definition point of view. Yes, I realize there is no way to gather all crazy ideas in one handy document, but some of the most frequent errors could be presented. Perhaps then we would not have debates like this one.

ISO did it in one of their standards - ISO 14405-2:2011 "Dimensions other than linear sizes". The one attached below shows ambiguity of directly toleranced dimension applied to stepped non-opposite surface (but the example of directly toleranced dimension applied "between" surface and virtual center of hole is also there).

http://files.engineering.com/getfil...b7-afd4-446abe061950&file=linear_distance.png

It would be quite informative for Y14.5 users to see something like this, don't you think?

Oh, and it would also force the committee to improve figures 2-4, 3-29, 8-7 from '09 edition

 

[fsincox]

I believe this is the same thing I was getting at. First you have to get the committee to make it clear the old way is not suitable and then you have to convince thousands of buisnesses that it makes a difference to them. It seems kind of like convincing people thay should invest in health insurance, when things are going good they don't want to listen, when things go wrong then they wish they had.
Frank

 

[greenimi]

Pmarc,

Just to blend a little my response with Frank’s, I would say that we are now in an intermediate phase of the profile “definition”.

1994: Fig 6-18 concept has not been included into 2009 (transition phase). I don’t know why, maybe because was never valid a truly design / possible real design requirement. This (transition phase) is the equivalent of “mandatory health insurance”, but if you don’t have it (insurance), you pay a small fine.

Now, maybe in 2012 or 2024, Y14.5 standard will no longer allow mixing of directly toleranced dimensions with profile, so Fig 8.27 / 2009 will be no longer carried over in 2021/2024 Y14.5 standard and accordingly the principle supported by the standard with Fig. 8.27 will go away. (why I said 2021 because 1994 - 1982= 12 + 2009 = 2021) along with improvements for the picture you mentioned (2-4, 3-29, 8-7). And to "help" Evan maybe 8-18 will get a revision too.
(and BTW the health insurance premium will be directly seen as direct deduction from the payroll same as the social security is now with no longer option to pay a fine and go without)

 

[pmarc]

greenimi,
I have no objections to mixing profile with directly toleranced dimensions in general. I just raise a red flag when it is done in a way that creates ambiguity.

There are quite few cases (per my experience) when mixing profile with directly toleranced dimensions does not lead to ambiguity, but there are such. Fig. 8-27 in Y14.5-2009 is one of them (although the very same requirement could be expressed through parallelism FCF with additional notation EACH ELEMENT below the FCF). But as we know, GD&T is a language, so the same thing can be said using different words/symbols. And I see nothing wrong with that as long as these different words/symbols are not in conflict with general rules of the language.

 

[fsincox]

pmarc,
Thank you for the link, I have added it to my arsenal to show it around here.
Frank

 

[pmarc]

You're welcome, Frank. Show it - in my opinion this picture is worth more than thousand words.

 

[fsincox]

It helps with Engineers, who do not understand all of this, if it has the "national standards" stamp and it gives it a "well this is what we have been missing" kind of feel.
P.S. You know how I like information on ISO standards
Frank

 

[pmarc]

I would say the content of ISO 14405-2:2011 is quite revolutionary. Y14.5 is not that far yet, but hopefully it will in 2021 or 2024.
It is really worth to have it on the shelf.