Datums and general GD&T assistance

[lowerp]

I have a customer with a plastic injection part design, similar in size and shape to that of a scotch magic tape dispenser, but with more bosses, ribs, and holes. This is a key component in a larger assembly.
They state tolerancing is per ISO 8015:2011 and ISO GPS. They have 7 different datums, 5 of them are axis of 5 different cylinders randomly located throughout the part (not a pattern). Datum A is the large flat surface, which makes sense to me for this component. The 5 different cylinders with datums are perpendicular to datum A, however, all degrees of freedom can be eliminated with datum A and the first 2 axis datums they identified. The customer explained to me that they added the additional axis datums in order for features referencing them (and closest by design) to have a better chance to meet position tolerances. This would involve numerous setups for inspection based on the feature control frame requirements.

My experience is typically with 3 maybe 4 datums, on smaller less complex injection molded parts.
Is it common practice to do this, almost treat the various areas of the part as independent from the other areas?
I don't see anything in the ASME standard that doesn't support this, so my assumption is ISO is ok too.
I did warn them that they may miss knowing what the correlation is between all features using just 1 datum simulator setup. I also suggested they consider 2 datum planes that run through the "middle" along X and Y. Any thoughts or suggestions would be appreciated.

 

[3DDave]

They cannot be completely independent, but they may have an extended chain of requirements if the sub-areas of the part have more-or-less independently operative features. For example: on a ship it may not matter where the winch hole pattern is at one end relative to another winch hole pattern at the other - they may independently be closely controlled to their nearby hatch locations. So, while the hatches need to be located to the deck outline, and therefore each other, each winch could have entirely different datum reference frames. This may repeat thousands of times in a ship.

So, is it common? Maybe not on a tiny part, but usually for lack of complexity and because CMM operators hate doing multiple setups and will refuse, through QA, to approve a drawing that does. It's a way of losing some useful parts in exchange for decreasing inspection time.

None cares, in the example, if the location of one winch hole pattern has shifted relative to another as long as each is in a useful place relative to the hatch - it may not matter in this case what one feature does relative to another if their functions are not closely tied either.

Or it may be that someone doesn't understand how feature control frames work. Don't know for your exact case.

 

[drawoh]

lowerp,

Normally, you need three datum features to immobilise your part. You may have the additional requirement to locate something to some non-primary feature, which requires a fourth datum feature.

--
JHG