GD&T question for 2 keyways on shaft

[potrero]

Please see attached drawing for details - comments are enabled in the PDF...

I've got a shaft with 2 keyways, which should be inline with each other along the shaft. My question relates to how to properly spec the drawing; is it sufficient to use "2X" and only detail one keyway, or is there a better way to detail this drawing? What is the appropriate way to indicate that I want the centerplanes of each keyway to be coplanar with each other (with a bit of tolerance)?

Thank you.

 

[drawoh]

potrero,

You called up one of your 1.75" <grin> diameters as datum_A. I do not think this datum scheme can be fixtured with any accuracy, especially at LMC.

I would have called up the other diameter as Datum_B, and used two datums to define your centre. The perpendular face you have as datum_B would become datum_C, and your slot side would be datum_D.

Your tolerance boxes would show datums |A-B|C|D|. In ASME Y14.5M-1994, look at figure 4-19. This is clearer than anything I can do with the text editor.

You can specify 2X on the features of size for your keyway. I would specify location tolerances separately on the length of your first keyway, and on the width and length of your second keyway.

JHG

 

[powerhound]

potrero,

Regarding the alignment of the slots, there is a "simultaneous requirements" rule in GD&T. You can use this tip for a similar example to what you are trying to do:

http://www.tec-ease.com/tips/january-00.htm

The important thing to remember is that both features must use the same datum reference frame. This does not mean they have to have the same tolerance in the instance that you invoke the rule on features of different size or features that can have different tolerances.


Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[potrero]

thanks to both of you. I've incorporated the A-B method, as well as the simultaneous requirements rule to handle alignment.

New rev is attached. What do you think?

Thanks.

 

[powerhound]

The only thing I would do differently is to change the dimension on the slot length from the ends instead of centerlines. As you now have it, the 20mm dimension is not a feature of size and thus, should not be located using position. Go ahead and make it 32.7 and put a 2X R on the end.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[potrero]

powerhound,
got it. on the 2X R... should it be 4X R in this case because there are 2 slots?
thank you.

 

[Xplicator]

Since this part is symmetrical, why is the shoulder on the left more important than the one on the right? This will create an R&R issue when inspecting. I don't believe it is, so a secondary mid-plane datum is preferred here. I also would consider use of composite position on the width of the key slots if you truly want to allow some amount of deviation between the two features.

Also, be very sure you understand the effects of zero tolerancing as well as including bonus and allowable datum shift before you so liberally apply them. Perhaps a tol stack is in need to justify.

This next suggestion is for all who will listen and comes from Lowell W. Foster himself, never, ever use dual unit dimensioning it just opens the door. Pick one and go with it. Remember it is a legal binding document. Lowell told me has found himself in many court cases testifying as an expert and this particular issue never bodes well for the supplier of the specification.

 

[potrero]

Xplicator,
Not sure I follow on your comment about the left shoulder being more imporant than the right. If I'm interpreting you correctly, then you are correct - what's important is that the key is centered between the shoulder "C" and the retaining ring groove. I dimensioned from "B" just to give a way of locating the keyway. How do you suggest dimensioning the keyway location?

Also, can you explain which feature would get the secondary midplane datum? (I think you mean that the dimension in zone E2 would get a midplane datum similar to how datum D is defined.)

Regarding zero tolerancing at MMC:
My goal is to have axes A and B, be coaxial. Then runout on the other diameters to be as specified. I want to have a spec which is not burdensome to inspect or produce; this is a machine shaft which will have sprockets mounted on the 50.8mm dia and bearings on the 44.45mm dia. Would it be better to spec the 44.45mm dia's with a small tolerance at MMC?

 

[drawoh]

potrero,

You do not need the run-out specifications on the 44.45mm diameters. These two features are in exactly the correct position. You have arbitrarily specified this by setting them as datums A and B. The only possible thing you might specify is roundness, and this only if your diameter specification does not control it enough.

JHG

 

[potrero]

jhg,
right... I don't think I have spec'd runout on the 44.45mm dia's...only on the 50.8mm dia's and the 57.15mm dia. I don't see a need for roundness spec. Not sure how to interpret your comment?
Thank you.

 

[weavedreamer]

As Datum D is not referenced in any of the other feature control frames . . . is it really necessary?

 

[Xplicator]

What I meant is that this is symmetrical part and that why should one shoulder be a datum over the other shoulder. So make you secondary datum a mid plane between the two shoulders using the 380.1 FOS as your datum feature. I would also however replace the runout with position on the two 50.8 DIA and the 57.15 DIA, it's more economical. If you want to check runout do so on the assembled product as a test procedure. As weavedreamer has stated unless you have other features with respect the DRF of A-B|C|D I see no need for you to call it out. Just use a position on the key slot pattern and if you want further control between the two slots make it a composite and refine the tolerance in the lower segment.

 

[drawoh]

Xplicator,

A mid-plane cannot be a datum. You can select either one of the two shoulders to use as a datum, or you can select the dimension across the two shoulders and specify the datum at MMC. I prefer it the way potrero has done it.

weavedreamer,

Datum_D is needed to locate the second slot.

potrero,

I said you could use the roundness specification if you wanted. It would allow you to control roundness more accurately than your diameter tolerance. Your diameter tolerances are tight, so your parts are quite round without an additional specification.

JHG

 

[Xplicator]

JHG
Are you kidding me? Better check again!

 

[Xplicator]

If I wasn't clear here it is again, The secondary Datum should be placed on the 380.1 WIDTH dimension FOS probably associated with a position w.r.t. A-B defining this as a mid plane datum and it is not necessary for it to be at MMC. This communicates the symmetry desired. And if you use position invoking BOUNDARY on the pattern of key slots, a tertiary is not necessary unless you need control of other features to the key slot.

 

[powerhound]

I agree with weavedreamer and Xplicator that datum D is not necessary unless you are trying to control something else besides the orientation of the slots to each other. The simultaneous requirements rule takes care of their orientation to each other and you don't have to invoke the BOUNDARY principle to invoke simultaneous requirements. What will have to happen though is that the callout for the slot length will have to be the FOS of the slot and not the centers of the radii.
There's another good tip this month on

http://www.tec-ease.com

about simultaneous requirements. I think it's February's tip.
One more suggestion is to add the standard to the note section to actually invoke it as the standard of interpretation.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[potrero]

Xplicator,
Is it really more economical to specify position on the two 50.8 DIA and the 57.15 DIA? According to one of the tec-ease tips (

http://www.tec-ease.com/tips/november-99.htm

) runout can be more economical to specify...

 

[potrero]

... meant to include this: this shaft is similar to the tec-ease example in the sense that it has a very large ratio of Length/Dia

 

[potrero]

The intent of this shaft is that there are sprockets mounting on each of the 50.8 dia's, and bearings on the 44.45 dia's. the 57.15 dia is a spacer between the sprockets. Would it be best to spec:
(a) the 44.45 dia's with position tolerance and assign datums A MMC - B MMC
(b) the 50.8 dia's with position tolerance, same as above, possibly with some amount of tolerance at MMC
(c) the 57.15 dia with runout w.r.t. A-B
?

 

[Xplicator]

If the only function of the 57.15 DIA is a spacer why do you care what the total runout is? What are the RPM's?

I guess it depends on who's perception you subscribe to on the economical part on position vs total runout. Alex Krulikowski seems to believe that it is and so do I.

Powerhound,
The invoking positional boundary is for slots that is why it was put in the standard see 5.10.1(c) regardless of simultaneous requirement. Again it is only a suggestion for consideration here as another method for a slot.