UF and collection planes in ISO GPS

From ISO 1660:2017:

Thank you pmarc.
That's is very helpful.
Should I understand that if the closed contour features are external then collection planes supplemental indicator is needed, but if the features are internal the collection planes is not "a must"?

There is no such rule.

The rule in ISO 1101 is that the collection plane indicator shall always be specified for all-around callouts in 3D definitions (MBD), but it's optional in 2D.

pmarc said:

The rule in ISO 1101 is that the collection plane indicator shall always be specified for all-around callouts in 3D definitions (MBD), but it's optional in 2D.

Click to expand...

If the model has associativity between the callout and the surfaces as required, with the visual response upon clicking and everything, what is the collection plane for?

That's a good question, which I deliberately did not want to introduce in my previous reply because I wanted it to stay focused on the original question.

I could have also probably said that for the vast majority of internal united features the usage of collection plane indicator adds very little value, as there is no ambiguity associated with the selection of the plane in which the all-around requirement is to work.

I give - how is an all-around internal feature unidentified "collection plane" not ambiguous when the inverse geometry of an external feature is ambiguous without one?

The main oddity I see in this is that the datum reference has been moved out of the whatever ISO calls a Feature Control Frame. The secondary oddity is that UF and all-around seem like they are surface set grouping operators and would tend to overlap in that capacity.

Is this just the Redundant Redundancy Bureau making a redundant marking to ensure that users aren't missing one of the redundant requirements?

3DDave said:

The secondary oddity is that UF and all-around seem like they are surface set grouping operators and would tend to overlap in that capacity.

Click to expand...

In ISO, all-around doesn't group the features, at least not in the sense of fixing the tolerance zones to each other. It only indicates which surfaces are controlled for by the tolerance, but without the datum-related constraints, the limiting zones can be disjointed (controlling each surface's form separately).

In 3D, for a simple rectangular block with a square hole passing thru a pair of opposed wall, applying profile tolerance all-around to the external shape without additional indication leaves room for ambiguity as to which 4 outer wallsof the block are controlled by that tolerance (of course if we forget that the object associativity exists and can be easily leveraged to eliminate the ambiguity).

However, for the square hole, I don't see a point in using the collection plane indicator in conjunction with the profile tolerance as there is just a single choice possible.

Neither the all around symbol nor the collection plane indicator is the tolerance zone grouping mechanism in ISO. In frequent cases, even the existence of datums may not be enough to group the tolerance zones entirely.

Also, regarding:

3DDave said:

The main oddity I see in this is that the datum reference has been moved out of the whatever ISO calls a Feature Control Frame.

Click to expand...

I guess you can call it an oddity, but the meaning is not the same as a datum reference inside the tolerance indicator (feature control frame). The profile zone in the example above is not enforced to be perpendicular to datum F. The collection plane is only used to identify the features that "all-around" applies to (the features intersected by a set of planes parallel to datum F).

Collection planes, intersection planes, direction features... sometimes even some of these combined... they are recent methods introduced as part of what seems to be policy to eliminate any view-dependent interpretations from ISO GPS. It's a solution that is worse than the problem.

"It only indicates which surfaces are controlled for by the tolerance" sounds like a grouping operator.

"the limiting zones can be disjointed (controlling each surface's form separately)" sounds like a random and not useful rule from ISO. Like, here are 4 contiguous surfaces, but "all around" means they can fly off into space regardless. Seems handy and would enjoy seeing an example where that was the preferred solution.

----

"as to which 4 outer walls of the block"

So, it could mean front, right, top and left sides? Sounds like Picasso has been doing profile portraits. What makes sense for that is to drop the word "profile" and replace it with "surface" so that it no longer carries the projection connotation that all other uses of the word "profile" have in geometry.

"Neither the all around symbol nor the collection plane indicator is the tolerance zone grouping mechanism in ISO. In frequent cases, even the existence of datums may not be enough to group the tolerance zones entirely."

That part makes sense. Because why, on a single part, would the surfaces of the part ever be held in relation to one another? That's completely insane. /s

Multiple surfaces tied together by nominal topology and basic dimensions are free to float away from one another? Does this in some way mimic the way an actual part, pulled from its box acts on the way into an assembly? If not why would one chose otherwise?

I thought Europe was a leader in removing lead pipe from the water supply.

Anyway, it appears that the obvious solution of using the word "group" and then the type of group would help, but maybe not. Everyone would then argue what the word "group" meant; probably deteriorating into features, surfaces, and derived geometry with no relation to each other at all until some more complicated phrasing was added.

"The surfaces identified as "A" (that is using the glyph that correspond to the "A" character as defined in the ASCII symbol list by placing the "A" glyph with a leader to the surface or terminating within the projected boundary of the surface in the view or directly on the surface as viewed in the model) are considered to be a group for topological purposes but not for GPS constraint purposes."

Click to expand...

Nope. Still too ambiguous. Needs to be longer.

Yep, that's the "independency principle" for you. This is how far they've taken it.

"The profile zone in the example above is not enforced to be perpendicular to datum F."

Ahh, it can be a millionth of a degree to datum F, smeared at an angle and millions of meters wide?

Is there an example of how that is done?

Could datum F be the bottom face and it's just the same expectation?

This is fascinating.

I understand the value in straightness of line element requirements where there would not be a datum reference; this is not a view dependent issue.

They must have eliminated the default perpendicularity expectation, if ISO ever had it, for basic dimensions. Like that [20], is that from the face or the edge or a vertex? It could be in any direction.

"Ahh, it can be a millionth of a degree to datum F, smeared at an angle and millions of meters wide?"

The tolerance zone can be fitted to contain the as-produced surfaces. it's not required to be 90° "theoretically exact" (basic) to datum F.

As for the "grouping" by the collection plane, it is just for identification purposes, based on the nominal geometry. It doesn't establish a geometric relationship between the tolerance zone and the datum, or between the surface elements forming that shape.

"this is not a view dependent issue."

The intent is that regardless of where you put the tolerance indicator - whether it's on the 2D view on the drawing, in the notes section of the drawing, on the 3D model, or on a supplementary document not depicting any part outlines, the all-around requirement is interpretable, that is - you would know "which way" is all-around.
I'm not saying it's a good idea (actually I think it sucks), but that's what ISO has now.

If it is unrelated to datum F and cannot be related to datum F then why is datum F mentioned at all?

As I said, "it is just for identification purposes, based on the nominal geometry". Meaning, any surface nominally intersected by any plane parallel to datum F is under the all-around control. However, that doesn't constrain any degrees of freedom or the tolerance zone to datum F. Messed up? Yep, it sure is.

Burunduk said:

.. they are recent methods introduced as part of what seems to be policy to eliminate any view-dependent interpretations from ISO GPS. It's a solution that is worse than the problem....

Click to expand...

Burunduk,

Why do you think that the cure is worse than the disease here?
I think, at least for the MBD perspective, is a good move to cut ties with the dependency view. You might have good reasons to think otherwise therefore I am trying to find your reasons, if you could please provide them here.

I guess, without putting words in your mouth, the complexity is one of the reasons?

pmarc said:

Neither the all around symbol nor the collection plane indicator is the tolerance zone grouping mechanism in ISO. In frequent cases, even the existence of datums may not be enough to group the tolerance zones entirely.

Click to expand...

pmarc,

So in ISO beside of CZ, SIM which are known to me (with my limited level of education in ISO hence questioning here) what are OTHER tolerance zone mechanism groupings?

And the second related question, if the datum system is locking all non redundant available degrees of freedom, based on your experience is a good idea (or bad idea) to still use CZ’s or SIM’s (just to make sure redundancy or lack of thereof does not get you in trouble)

greenimi,
Here is why I think the 'collection plane' cure is worse than the disease.

First, it is redundant—especially for MBD. MBD standards such as ISO 16792 and ASME Y14.41 require geometric tolerance annotations to be associated with the features they control. For example, when you click on the feature control frame, the associated surfaces are highlighted. On 2D drawings, it is also redundant. The view in which an all-around callout appears makes it clear what it applies "around".

Secondly, the callout is long and, as you mentioned, complex. As you can see from this discussion, the unusual use of a datum reference causes confusion about what the datum does. Which means it can easily be misinterpreted.

If the front surface is not parallel to F then it is also collected?