Hole Pattern as Datum Feature at LMB

[gtadams]

Hi all,

I'm looking for some guidance on a proposed tolerance scheme that uses a 2 hole pattern as a single datum and is referenced at LMB. As for as I can tell, this is legal per the 2009 standard, though maybe inadvisable for high volume production since it requires a CMM instead of a functional gage.

Attached is a quick sketch showing what I'm trying to accomplish and the proposed method. The tolerances are just arbitrarily chosen for the example, but the question is focused on the usage of the datum features.

Is this legal, and is it actually the best way of accomplishing my alignment goal?

Thanks

 

[Burunduk]

It says: "Goal: control the alignment of these cylinders as closely as possible". Sounds like a case for referencing the datum features RMB.
LMB permits datum shift when the datum features depart from the LMB boundary, and datum shift in both parts can contribute to misalignment of the considered features.

 

[drawoh]

gtadams,

Your design does not work. Your positional tolerance must not exceed 1/2 the hole clearance. You have not specified your pin diameters, but your hole tolerances are very accurate. If you are using two pins to locate a part, one of the four holes needs to be a slot.

The alternate approach is to specify hole clearances that are double your positional tolerances. If your positional tolerances are achievable, the error between your two cylinders.

Note how positional tolerances are based on minimum clearances, i.e. everything at MMC.

--
JHG

 

[3DDave]

drawoh - one of the pins is in a cross-wise slot in the base.

 

[gtadams]

Burunduk,

If I choose RMB, won't my worst case alignment (when my parts are at LMC) be the same as if I used an LMB modifier? For all parts other than the ones that end up at LMC, I agree I would get better alignment. But if I must be prepared to accept parts at LMC why would I not allow the bonus from LMB?

 

[Burunduk]

gtadams,
When you specify LMB, the worst-case alignment can occur when the datum features are produced at MMC, because that's when you get the maximum datum shift. That's when the tolerance zone for the cylinder in each part can shift up to some amount of displacement relative to the datum features (because the datum features can shift relative to the datums).
With RMB, you get no datum shift at all. The alignment will depend only on the position tolerance and the maximum clearance between the parts.

 

[gtadams]

I feel like I must be missing something here. If I specify RMB and my parts are at LMC, isn't my alignment the same as if I specify LMB and my parts are at LMC (no datum shift)? And then similarly, if I specify LMB, isn't my alignment the same between LMC with no datum shift vs MMC with a datum shift?

 

[Burunduk]

If I specify RMB and my parts are at LMC, isn't my alignment the same as if I specify LMB and my parts are at LMC (no datum shift)?

Let's start from this:
"No datum shift" when the datum features are referenced at LMB and produced at LMC is not necessarily true and it's not so in your case. For example, for the position on the red shaft, the LMB boundary of each of the two holes that are part of datum feature E equals 3.26 (because that's their OB with reference to the primary datum D as imposed by the position tolerance applied to them), whereas their LMC size is 3.01 each, so that's up to 0.25 available datum shift when any of them is produced at LMC.

Regarding the alignment comparisons between the different conditions that you ask about, I still tend to say that the worst alignment when the datum features at RMB is better than the worst-case alignment with a material boundary modifier because generally, material boundary modifiers allow for additional variation. But I will leave it to those that are good at stacks and analyzing these kinds of things to say the final word. Hopefully, you will get a well-supported answer.

 

[gtadams]

Ah, I was completely neglecting that LMB includes the position tolerance of datum E, so now I see how RMB is the right choice. Thank you for that.

The follow on question then, is there any way of taking advantage of the size of Datum E when it's not at LMC? I could theoretically allow a larger positional tolerance on my feature as datum E departs from LMC on the basis that the worst-case alignment is the same. But the only way I saw to do that was to reference E at LMB, which you pointed out incurs that 0.25 penalty even at LMC.

 

[Burunduk]

I could theoretically allow a larger positional tolerance on my feature as datum E departs from LMC on the basis that the worst-case alignment is the same.

Not sure about that. Your datum feature E are two holes into which dowel pins should fit snug at any allowed size of the holes. The one pin that assembles into the hole at the blue part will be the locating component so the size & orientation of that pin is what matters. Is assembly clearance a good thing or bad thing in your case? Can you use the clearance between parts to adjust the alignment manually before clamping the red part in place by some fastener not shown in your image or will it only cause instability/uncertainty?