HOLE AS PRIMARY DATUM OR NO?

[durablack2]

I have a part that is stamped from sheet metal. There are about 6 rivet holes that should be referenced off 1 larger hole for a pin that is about 25mm in diameter. I have made an example of a similiar part that displays the several methods for identifying the datums on this part. The edges of the part do not really matter that much so I know these should not be the primary datum. Should the 25mm hole be primary datum and bottom surface of sheet metal be secondary? Or is it best practice to make the bottom of a sheet metal part the primary datum, and call out the 25mm hole as secondary datum, and assign perpendicularit? OR is it better to use the bottom plate of sheet metal as the secondary datum and call out perpendicularity to axis of 25mm hole?

Which, if any, are correct?

 

[TheTick]

I've never done it. I've seen it a lot of places. I've never had cause to call out position with less than two datums.

 

[Belanger]

Tick -- position's main job in life is to control location. So if we try to use position on a single hole with a single datum reference (a plane perpendicular to that hole), then it's not even doing its intrinsic job (location).
It would be like applying parallelism on a flat surface with no datum reference, and saying that it's identical to flatness -- it would be failing to do its main job (orientation).

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

 

[3DDave]

Just because I've been here before - isn't position's first job to see that all the multitude of elements of a feature are within a certain mutually shared zone? If they are not it doesn't matter where the zone happens to be as the item will be noncompliant.

This use is seen when two nominally coaxial holes are held to a mutual position tolerance without reference to any datum at all. Since this usurps using straightness of a potentially continuous feature, is this use also to be rejected? Likewise if two parallel holes are defined with a single position tolerance how is the inspector expected to fish out where the zones are relative to the part? Would they not line up one hole and then change the part orientation until the second one was in an alignment the CMM would find acceptable or the second pin on a fixture lined up?

This seems more like a convention than an impossible case. Is this more about keeping the standalone orientation tolerances that might otherwise be replaced?

Perhaps it's time for a refactoring of the concepts to obsolete the artificial subdivisions?

 

[pylfrm]

Perhaps it's time for a refactoring of the concepts to obsolete the artificial subdivisions?

Any maybe shave off a hundred pages or so? If only...


pylfrm

 

[Belanger]

isn't position's first job to see that all the multitude of elements of a feature are within a certain mutually shared zone?

No, this isn't position's job at all. Position is looking at the axis of the UAME, so it's not controlling straightness.

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

 

[Sem_D220]

Since the OP noted that the 4 end faces of the part do not matter and if I understand it right - not functional for assembly, I thought of proposing another scheme based more on the holes constructing the DRF.

What are the opinions on the following:

Primary datum feature A - bottom surface of sheet metal.

Secondary datum feature B the large hole.

Tetriary datum feature C - one small hole, with the note "INDIVIDUALLY" next to the datum feature symbol.
Edit: after rethinking, "INDIVIDUALLY" won't be needed, as the datum feature symbol will have a direct leader to a specific hole, and placed separately from the pattern position FCF.


The 2 small holes as a pattern ("2X") positioned wrt DRF A|B|C, datums B and C called out RMB, or MMB, whatever is more appropriate functionally.
Will anyone consider this self-referencing?

The four end faces of the plate can be controlled with a sloppy enough all around profile to the same DRF.

semiond