Compound datum feature A(M)-B(M) used as a primary

[greenimi]

A cylinder has three inside diameters,

ID#1 on one end ½-20UNF thread (called datum feature A)
ID#2 middle section ø.375 and
the other end ID#3 again ½-20UNF thread (called datum feature B)

Datum feature A is ID#1 thread
Datum feature B is ID#3 thread

The middle section width/length is dimensioned with ± and is called datum feature C
Other features on this part (such as holes perpendicular to axis of the cylinder) are positioned to A(M) –B(M) primary and C (RFS) secondary.
The two inside mating parts (one on each end of the cylinder) threaded on ½ -20 thread will create a sealing surface/contact with datum feature C (and that’s why datum feature C is called RFS). In other words, take the mating part and threaded all the way in until touches one side on datum feature C to create a contact surface. The same thing on the other side (just to describe a little bit the functionality)

Is this compound datum (used for threads !!) a good option? I would say it is legal option, but it is a functional one?

 

[greenimi]

pmarc,

Did I answer all of your questions? Does it make any sense how this assembly suppose/ is intended to work?

Thank you for your help

 

[pmarc]

Actually I still have some doubts.

Allow me to express just one of them in a following way:
Based on the description you provided (that both plugs locate and orient the green part in the assembly, and they are able to move radially relative to each other to some extent), do you think it is functionally justified to have threaded datum feature simulators A and B (used to establish a single datum axis during inspection) fixed in location relative to each other (as this is what default rule says)?

 

[greenimi]

I don't really know if it is functionally justified, recomanded or even feasible and that IS the main question of this thread.

But, why do you think that the datum feature simulators A and B must be fixed in location relative to each other? Or do you think is better to use translation modifier (as per Y14.5-2009)?
Again, this is what the current print said before someone start questioning DRF.

 

[pmarc]

Yes, I am thinking of using translation modifier to untie default 0 basic linear relationship between datum feature simulators A and B. Why? Because in general the method of setting up a part for measurements should reflect real function of the part in an assembly. If you are saying that a mutual radial movement of both plugs is possible and that the plugs orient and locate the green part, this should be taken into account by inspection. Otherwise the part may be inspected relative to a DRF derived from too restrictive, not 100% functional, configuration of datum feature simulators.

Going further, perhaps it would even make sense to untie orientational relationship between datum feature simulators A and B (within defined datumless positional tolerance applied to both threads)? The question is: do both plugs "behave" in the assembly that way? The problem is - how to say that in GD&T language known from Y14.5-2009.