Slot question

[cjccmc]

We use Model Based Definition(MBD)at my work which essentially means you query the cad model to get dimensions. I used the TOP method per Y14.5M-1994 for slots(page 143 or so I believe) giving about 3X the tolerance along the length as that on the width. Now I'm wondering what if anything controls the tangency of the end radii to the width. Being MBD there is no "2XR" label on the end radii. Theoretically, can the part be produced with an abrupt mismatch between the radii and the flats of the slot?

 

[ctopher]

I call out "R FULL". This makes the radii tangent to the straight width.
But, the machinist can still make a mismatch.

Chris
SolidWorks 11
ctopher's home
SolidWorks Legion

 

[KENAT]

You've found one of the limits of MBD as often implemented. It actually leaves a lot of i's undotted and t's uncrossed if you're not careful. Y14.41 is of more use for CAD companies than MBD users last I looked.

Is your your approach to MBD that the model is 'basic' and you have some kind of default surface profile tolerance for all surfaces/features not directly dimensioned?

In this case I'd think the general 'form' of the radii is controlled by the surface profile tolerance. However, I'm not sure this is explicitly supported by the standard, at least 14.5-1994.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[cjccmc]

Yes, queried dimensions from the model are basic. We do have a default note:

PROFILE TOLERANCE FOR PART AND ALL PART FEATURES, EXCEPT FOR HOLES, IS +/- .020 UNLESS OTHERWISE NOTED.

But I think that the TOP on the slot qualifies as the "UNLESS OTHERWISE NOTED".

 

[KENAT]

First off the way you spec your default profile tolerance is odd and doesn't really comply with my understanding of 14.5. Profile is specified as a total allowance - default equal bilateral - not +-. There are ways to conventionally (as in 2D annotation) indicate unequal bilateral or unilateral (fig 6.11 in 94 version) if needed.

Now on your actual question on the slot, yes as is it's ambiguous. To try to apply surface profile to the radial portions while simultaneously having +- size limits on length and width and true position controls on location in the 2 axis does seem problematic - that's what I meant by not being sure these overlapping requirements are supported by the standard. A simple fix/interpretation doesn't spring to mind - sorry.

Or are you saying you don't have +- on the actual length/width - just the position tols?

As to specifically ensuring the radius is tangential, I don't believe 'R' on a drawing explicitly achieves ths - to do so required 'CR' (section 2.15).

I'm trying to think of a way you can achieve what you want by having relatively tight surface profile control on the straight portions and much looser surface profile on the radii. However, to get your tangency you'd still need to specify 'CR' which almost brings us back to the initial issue.

Can you really not add a "2X R" callout?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[cjccmc]

Thanks for your replies.

I think we arrived at the "+/-" on the profile note because we kept getting inquiries ("does that mean .04 in each direction?"). It seemed to clarify by saying +/- .02 as most understand it is .02 in both directions even tho it conflicts a little with the std as you point out.

"Or are you saying you don't have +- on the actual length/width - just the position tols?" No, I do have +- on length and width.

I think I'm over-thinking this slot tangency issue. There may be a theoretical problem but I've never seen any parts made with bad mismatch at the radii. The more I think about it there seem to be many (theoretical) ways to produce a part that meets all the literal tolerance requirements but would be butt ugly and problematic. Good thing most machinists understand and provide the added niceties that are expected by the end users.