Profile tolerance unless otherwise specified - WRT Datums vs no datums 1

[Frokilin]

Hi,

I have read some post but I didn't fin a clear answer for this question. When you have a drawing with datums defined, and a general note saying "Unless otherwise specified profile tolerance of: "
What is the difference between having that surface profile called out back to datums or called out to no datums?
Besides having those surfaces better controlled, is there any other reason?
Thanks

 

[pmarc]

It's not arbitrary; one simply has to look at how the surface (well, line) is related to the datum. And in that case it's not a basic dimension, so we are to understand that it's doing the lower-level control of orientation. That's the specific indication that you're looking for.

I used this example in some past discussions in the forum, so I'll use it again here. What if the surface with the line profile callout was nominally coplanar with datum feature A? Would then a directly toleranced zero dimension need to be applied to accomplish the similar effect as in fig. 11-32? Why wouldn't someone "just" use the approach shown in fig. 11-31 instead, where the line profile FCF explicitly informs everyone, including software, without the need to search all over the drawing for the related +/- dimension, that the datum feature references shall only constrain the rotational degrees of freedom of the line profile tolerance zones? Why couldn't the same approach be used in fig. 11-32 to avoid the controversy?

Garland23: I'm sure that you agree that if the symbol in Fig. 11-32 were profile of a surface, it would be identical to using the angularity symbol.

If the symbol in Fig. 11-32 were profile of a surface, I would strongly recommend changing it to parallelism to avoid the exact conversation we are having now. As a matter of fact, this (the use of the parallelism symbol instead of profile of a surface) would be the direct indication that the FCF is not to control location and I doubt that anyone, including software again, would have any issues with that.

 

[Burunduk]

I'll disagree a bit. Paragraph 3.68 defines true profile as requiring basic dims to define the profiled feature itself. It does not mention anything about the feature's relationship to any datums.
If a DRF is imposed but the location dimension to the profiled feature is not basic, then it controls orientation (with the angle being implied basic).

There is actually no disagreement.
I was answering on OP's question "Where in the 2018 standard implies that the surface profile controls location only when there is a basic dimension?", and that's what the true profile definition quote was for. You can have an un-located true profile if that is what you need. But if you want a located true profile, you need basic location (on-drawing or from-model) and the basic location dimensions would be part of the true profile definition just like the locating dimensions of a true position.

And I think that this state of affairs is what provided the possibility to create the 11-32 example, which I personally can't stand. I think it doesn't provide any useful ideas. It would be better to specify parallelism or angularity on EACH LINE ELEMENT note basis relative to A,B than a crippled profile. I view profile as first and foremost a location control for surfaces because that's the only direct and meaningful way to locate surfaces (non-size features). Using it like they did in 11-32 is like using position to control the perpendicularity of a single hole to the surface it's designed normal to. Also, there should be no functional reason to do what 11-32 shows. The directly toleranced height suggests some kind of fit requirement with the mating part, which normally implies that the 2 sides of the regular FOS have equal significance and no sensible reason to use one of the sides as a datum feature for controlling the orientation of the other.