Primary datum A or B-C (primary)

[greenimi]

What a day……. I had a big argument with some of “local” GD and T “experts” and I would like to run this issue past you guys to see what you have to say.
I wouldn’t expect to have arguments about such a simple part, but anyway….
The part is “sandwich between” other two components called mating parts (mating part 1 and mating part 2). The requirement is to have the hole in the middle on the .500 width.
Now, we had position Ø.005 wrt A (no bonus and no datum shift allowed).
Some guys suggest having the position callout changed to position Ø.005 B-C (as primary), where B is the datum feature plane (.750x 1.000) on one side and C is the datum feature plane (.750x1.000) on the other side.
My opinion was that those callouts are equivalent and produce similar results/have the same effect. They are just two different syntaxes to say exactly the same thing.
Others did not agree and said that it’s a difference in the way the datum feature simulators act and because of that it makes the callouts different. On the position only to A the datum feature simulators is allowed to rotate until get the highest points on the flats, but on B-C (primary) the simulators must remain parallel to each other (they invoked some rule of the datum feature simulators such as must have perfect form, perfect orientation to one another and be perfect positioned) ---
4.5.2 -2009 page 53
I am not saying the rule does not exist, but I don’t know if it’s applicable in our case or not, and here I need your help a little bit.
Which position would you consider to be the correct one to be used in this case?

Thank you for your help

 

[CheckerHater]

I am simple person.
To me if your part is allowed to “rattle” between the datum simulators, then yes, indeed you introduce datum shift.
And then you say you do it to eliminate datum shift.
I would stop right there.
It looks like you have created MMD datum without using letter “M” on your drawing. Why?
I was honestly trying to understand what is going on and I think second option (B-C) should not be used just because of the level of confusion it creates.
Just an opinion.

 

[pmarc]

To be honest, I do not understand why some of you keep claiming that in B-C case both datum feature simulators are spaced basic .500 apart (or actually I understand why, but in my opinion this is not how it should be interpreted).

Short exercise:
I assume that current .500 dimension has a tolerance - probably defined somewhere close to a title block. Let it be +/-.010 for the purpose of the exercise. That gives us .490-.510 limits of size. Now, what if I expressed this dimension directly on the print area as .495 +.015/-.005 for instance. Would you still consider the simulators B & C as spaced .500 apart? Basing on what rule?

 

[Belanger]

Pmarc -- they are caught up in this notion of adding a basic dimension along with profile tolerances. But if the first option in the OP were simply embraced, this whole mess can be avoided. Perhaps it's just an academic discussion at this point.

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

http://www.gdtseminars.com

 

[flash3780]

I think that referencing a position relative to B-C is unconventional and confusing. If a centralizing datum is the intent, I recommend referencing datum A which should be well-understood.

I'm not sure of the function of the part, but you can probably increase your tolerance without negatively impacting the function of the part by adding a maximum material condition modifier to the datum reference (A) and to the positional tolerance. This will also allow the position of the features to be checked with hard gaging.

Also, you're not currently controlling the perpendicularity or the left-to-right position of the hole... I think you'll want to reference three datums to establish the positional tolerance.

 

[flash3780]

To clarify the above, I modified your sketch with my recommendation (see the link below).

 

[bxbzq]

pmarc,

Isn't para 4.5.2 in '09 std quite clear? In your exercise, I don't know how the datum feature simulators would behave. They would not be spaced .500 basic apart, they would not be movable within the range of width tolerance either in my opinion. Not defined to my interpretation.

Question to all, if the .500 was basic, and Profile tolerance was applied, would the height still be considered as FOS?

 

[Belanger]

To the last question -- no it wouldn't, because the standard defines a FOS as something that associated with a "directly toleranced" dimension, and a basic dim doesn't seem to meet that criterion (see paragraph 1.3.32).

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

http://www.gdtseminars.com

 

[flash3780]

I'm not sure of the advantage of using a profile control on this part, but even if a dimension is basic, it can still be a size dimension (and thus can define a feature of size). Ref. ASME Y14.5-1994, 1.3.17 (Feature of size definition) and Para 6.5 (Basic size dimension/profile control).

 

[Belanger]

So now the FOS question comes down to which standard you're using; the current standard added the requirement that it be directly toleranced.

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

http://www.gdtseminars.com

 

[pmarc]

bxbzq,
So are you saying that if the dimension in my example was specified as .500+/-.010 everything would be nice and clear? Or since the dimension in OP's picture is specified just as a stand-alone .500 everything is nice and clear?

As for profile tolerance application and dim .500 being basic -- this is totally different story. Look at fig. 4-33 in Y14.5-2009. The basic height in OP's case could be assigned as datum feature, just as in 4-33, and the datum would always be a center plane. In addition there would have to be two profile callouts (or one with two leaders) without any datum references applied to both surfaces of the height. And now different scenarios could take place depending on how datum feature A was referenced in positional FCF for the hole in the center:
- A without any modifier - datum center plane simulated at RMB. (So, in fact, this would be more or less equal to case A in OP's question);
- A followed by (M) modifier - datum center plane derived from height's MMB;
- A followed by (L) modifier - datum center plane derived from height's LMB;
- A followed by [BSC] modifier - datum center plane derived from basic .500 height;
- A followed by [VALUE] modifier - datum center plane derived from VALUE height.

 

[KENAT]

Greenimi stated at 22 Apr 13 15:19 what version they were using -1994.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[bxbzq]

pmarc,

No, I'm saying if .500 basic and reference B-C in FCF, it is clear the simulator B and C would be spaced .500 apart. Para 4.5.2 backs me up.

In your exercise, however, it is not clear how the datum feature simulator B and C would behave. Because the distance between surface B and C is not basic dimension. Would you agree?

To go on my question, as J-P pointed out, if the height is not FOS, why does the datum of B-C have to be the center plane?

 

[pmarc]

Yes, I would agree. And this was exactly my point. Without having basic orientational and/or locational relationship between B & C defined on the print one can not say that the simulators are spaced basic .500 apart. Moreover, one can not even say that the simulators are parallel to each other.

As for your question, I was attempting to say that if the HEIGHT was assigned as a single datum feature, like in fig. 4-33, and not surfaces B and C separately, there could be a couple of extra possibilities fully compliant with the standard.
I do not see however any value of having B & C specified as datum features separately, even with basic relationship defined between them.