Positional tolerance of a hole when datum is at an angle? 2

[rwild]

Hello all,

I need some help on dimensioning and inspecting a hole location with postional tolerance. Please see drawing. First is the proper use of GD&T, and if it is, how would we measure from datum B? I understand the concept of true position when the datums are all perpendicular to the feature, but the angle has me boggled. Thanks for your help.

Randy

 

[CheckerHater]

Something like this:

 

[pmarc]

rwild,
Look at attached snapshot. It is from Y14.5-2009 and shows how this should be done properly, and what is interpretation of datum reference frame.

http://files.engineering.com/getfil...-6e0d-45fe-99c2-4fdedbfd99db&file=fig_4_7.JPG

Notice that - apart from locating the hole relative to referenced datums - each datum feature of lower precedence is geometrically controlled to datums of higher importance. In other words B is controlled to A, and C is controlled to A & B.

 

[cjccmc]

Wondering why in PMARC's figure an angularity was used for datum B surface. Since its perpendicular to A why not perpendicular symbol?

 

[pmarc]

cjccmc,
Since 2009 edition of Y14.5 this is acceptable practice. After all, perpendicularity is a special case of angularity.
For more details see fig. 6-4 and paragraph 6.6.

 

[CheckerHater]

apart from locating the hole relative to referenced datums - each datum feature of lower precedence is geometrically controlled to datums of higher importance. In other words B is controlled to A, and C is controlled to A & B

Which always appeared to me as being optional

 

[pmarc]

Which always appeared to me as being optional

But not for authors of Y14.5 clearly stating in paragraph 4.9: "Geometric tolerances related to a datum reference frame do not take into account any variations in form, orientation, or location of the datum features. Datum features shall be controlled directly by applying appropriate geometric tolerances or indirectly by dimensions such as the size of a primary datum feature of size."

 

[rwild]

Thanks alot to all of you. This was very helpful.
Randy

 

[CheckerHater]

Pmarc,
Thank you. I am not sure the paragraph was there in 1994.
That brings interesting question though: if I must (or “shall”) apply geometric controls to my datums in order of precedence, what will happen if I want to use same datums in different order of precedence on the same drawing?

 

[pmarc]

CH,
Your interesting question brings another question - what is a functional need behind having the same datums stated in different precedence in different FCF's on the same drawing. I am afraid that imagining a part functioning in a way that such play with datum references would be required/justified is way beyond my

 

[pmarc]

abilities

 

[CheckerHater]

You cannot imagine part like this?

 

[pmarc]

Apparently my imagination is not working as well as yours.

 

[fsincox]

That statement was certainly was not there in 1982 or 1973. Thus the reason for some of the discussions PMARC and I have had.
ANSI Y14.5-1982 4.2.3 says:
"IF not sufficiently accurate, datum features MAY need to be controlled by specifying appropriate geometric tolerances"
Frank

 

[bxbzq]

Please go on.

 

[pmarc]

I think that CH's example may fall into category of "exception to the principle". The principle is stated in subparagraphs (a) through (c) of 4.9 in Y14.5-2009. The exception is that it seems technically impossible to follow that principle in the light of functional requirements shown.

But I think there is agreement here that datum features must be controlled to each other somehow. So if I had to choose something, I would still pick one datum configuration (e.g. A|B|C) and control other datum features according to the rules stated in 4.9.

 

[fsincox]

I do agree, I am just saying the committee has not always been quite so bold. I would also say, at best, this earlier philosophy left many with a mistaken impression which led to things like my "keyway" example.

 

[CheckerHater]

Apparently anything more complicated than flat rectangular plate with four holes in it is an “exception”.
Let say, I limit myself to just A|B|C in my example. That will create “simultaneous requirement” and fix everything nicely in space. Problem solved?
Unfortunately it will raise another interesting question: What if I want to refine my position and apply composite tolerance?
According to The Book my FRTZF has to follow the same datum framework as PLTZF.
I wouldn’t dare to interpret the meaning of all three mutually perpendicular patterns being “refined” wrt |A|. It just won’t make any sense, would it?
Don’t get me wrong, I agree that datum features are usually important for the part; I just prefer to control them according to functional requirements and not to be told what order I “shall” follow.
My interpretation of The Paragraph would be “You SHALL NOT expect datum features to control themselves by magic.” But I would stop there.

 

[pmarc]

Let say, I limit myself to just A|B|C in my example. That will create “simultaneous requirement” and fix everything nicely in space. Problem solved?

Partially. 4.9 could be used without any addtional caution, but on the other hand this would change functional requirements for 2 patterns that weren't initially controlled to A primary.

Let say, I limit myself to just A|B|C in my example. That will create “simultaneous requirement” and fix everything nicely in space. Problem solved?

I would say 3 composite positional tolerances with A in FRTZF would make sense - at least geometrically. Even A|B in FRTZF would work. Again, whether these would be reasonable choices from functional point of view is completely different story.

 

[pmarc]

This is how the second quote in my recent reply should look like:

According to The Book my FRTZF has to follow the same datum framework as PLTZF.
I wouldn’t dare to interpret the meaning of all three mutually perpendicular patterns being “refined” wrt |A|. It just won’t make any sense, would it?

Too much of copy and paste. Sorry for that.