Surface Profile Tolerance Without Datums - is it flatness? 2

[Radius1]

Applying a surface tolerance on a surface without datum reference. Is this flatness? My knowledge is limited in GD&T.

Is this true ? I am having a hard time seeing it.

I want to ensure that the profile of the surface is 1mm. I did not want to reference datums, as I read datums were not required for profile tolerances.

 

[Burunduk]

greenimi,
A valid angularity requirement (with datum references) applied to a surface also controls the form of the surface. In other words, the orientation control (angularity) also encompasses the maximum flatness error that will still keep the surface acceptable.

 

[pmarc]

greenimi,

It is correct that today orientation tolerances without a datum reference are not supported by Y14.5. I remember this idea (datumless orientation tolerances) was brought up to 14.5 committee's attention a few years ago but I am not sure if it has gained any traction since then. Personally, I would find it potentially useful is certain applications, although I think that in some instances a datumless profile can be used instead to achieve the same end result.

As far as how ISO handles the concept, ISO 1101:2017 in Table 2 states that orientation tolerances need datums. However, ISO also has the CZR (Combined Zone Rotational only) modifier which, when specified in a datumless profile tolerance, should be capable of controlling orientation between the features (tolerance zones) but not location.

 

[Burunduk]

ISO also has the CZR (Combined Zone Rotational only) modifier which, when specified in a datumless profile tolerance, should be capable of controlling orientation between the features (tolerance zones) but not location.

How does it differ for a simple datumless profile callout that applies to two non-parallel surfaces? Each tolerance zone can be extended as far as necessary until they intersect. What distance defines the mutual location between two non-parallel faces of a triangular prism?.
If the corner is trimmed by some other feature that doesn't really change much.

 

[3DDave]

Dimension [H] seems irrelevant as nothing locates its action horizontally.

Every unbounded taper has a range of possible heights from 0 to infinity so any similar dimension adds no actionable information.

The lesson seems to be that a surface with unclear bounds has unclear bounds.

 

[Burunduk]

Dimension H is irrelevant because two non-parallel surfaces don't have a mutual location relationship.

 

[3DDave]

They have an intersection point, locating the tolerance zone - exactly what your picture shows. The dimension doesn't locate the boundary of the surface that fits within the tolerance zone; the bounds are undefined.

Had the vertical surface on the left end also been given the same profile tolerance, then that boundary would be defined and [H] would become relevant and establishing the mutual location of that boundary in spite of the surfaces being non-parallel.

 

[Burunduk]

Had the vertical surface on the left end also been given the same profile tolerance, then that boundary would be defined and [H] would become relevant and establishing the mutual location of that boundary in spite of the surfaces being non-parallel.

That is correct.
Alternatively, if the vertical surface on the left had been used as a secondary datum feature in to establish the profile's datum reference frame, that location would also be defined relative to the established origin and 'H' would be meaningful.
But in the original context of the datumless two-surfaces profile tolerance (mentioned by pmarc), the question remains - why would ISO's CZR modifier be required? Unless the nominal orientation relationship is parallelism, how modifying the profile tolerance should change anything?

 

[3DDave]

I know that is correct. The question was about controlling orientation, not location.

 

[greenimi]

And one more direction for this thread:

If I change the H dimension from basic to plus-minus (±), keep S° angle “as-is” meaning basic then how this will affect the meaning of the callout (profile or angularity for that matter) and the end geometry or result of the specified surface?
Basically, I am asking how far from 0° zero basic or from 180° basic we can go and the callout still be meaningful?
I am asking this in the spirit of figure 11-32 -Figure 11-32 Profile of a Line and Size Control- from ASME Y14.5-2018.

If zero basic works than what about 1° basic? 2° basic?
Where we draw the line of our assumptions?

 

[pmarc]

In ISO, in cases where there is no default location relationship between features, CZR does no have to be specified, but still CZ is required to tie the multiple tolerance zones together.

The ISO part of my previous reply to greenimi was purely about cases where the usage of CZR would make sense (e.g., for multiple coplanar or parallel surfaces).

 

[Burunduk]

The ISO part of my previous reply to greenimi was purely about cases where the usage of CZR would make sense (e.g., for multiple coplanar or parallel surfaces).

This answers my question clearly. Thank you.

 

[Burunduk]

If I change the H dimension from basic to plus-minus (±), keep S° angle “as-is” meaning basic then how this will affect the meaning of the callout (profile or angularity for that matter) and the end geometry or result of the specified surface?

It will change nothing. In all cases except for basic distance which will also mean basic (implied) 0°, the location won't be controlled. Profile of a surface with ref. to A will act as angularity.

 

[greenimi]

It will change nothing. In all cases except for basic distance which will also mean basic (implied) 0°, the location won't be controlled. Profile of a surface with ref. to A will act as angularity.

So the fact H is basic or H is ± is irrelevant for the profile callout?
I am trying to get a clear understanding when the surface (the surface shown with profile callout in Burunduk’s posted picture) is controlled for orientation only and when the surface is controlled for orientation and location.

 

[Burunduk]

So the fact H is basic or H is ± is irrelevant for the profile callout?

In this case, it is so.
The sloped surface doesn't have a direct location relationship with datum A.
A location relationship can be established only through a third party surface such as the left side surface which has an intersection with the top face and that intersection is at a certain distance from datum A.
If the left side surface is used as a secondary datum feature for the profile control or take part in a simultaneous requirement relative to A, only then you can say that the H distance is limited by the specified tolerance or tolerances.

 

[greenimi]

The sloped surface doesn't have a direct location relationship with datum A.
A location relationship can be established only through a third party surface such as the left side surface which has an intersection with the top face and that intersection is at a certain distance from datum A.
If the left side surface is used as a secondary datum feature for the profile control or take part in a simultaneous requirement relative to A, only then you can say that the H distance is limited by the specified tolerance or tolerances.

Ok. Understood.
But if the S° angle is close to zero (and zero basic is implied and consequently not shown) then I guess H will be relevant, doesn't it?

 

[Burunduk]

But if the S° angle is close to zero (and zero basic is implied and consequently not shown) then I guess H will be relevant, doesn't it?

If there is a basic "H' distance on the drawing from the datum feature to the entire top face (and hence an implied zero basic angle), then the true profile and the tolerance zone defined by it are parallel to and located from the datum. So yes, in such case H is relevant, including when the actual surface is not perfectly parallel to the datum, which will always be the case. In such case, H defines the basic location of the true profile and the tolerance zone.
But if the basic angle is non-zero, and even if it's a small angle of less than 1°, then there is no longer a direct basic location relationship between the feature and the datum.