E-clips on centered shaft

[ant77]

Hi all. I am new to GD&T and need your help setting up a simple drawing for a shaft.

The 6.0mm diameter shaft is held in an extruded C-section by two e-clips, both inside the 'C'. As a result, the e-clip grooves are symmetrical about the centre of the extrusion.

Two other e-clips lie outside the 'C' section, and are also symmetrical about the centre of the section. One end of the shaft extends further from the centre of the channel than the other, in order for us to fix a pulley to it.

In the past I have drawn a center line on the shaft where the center of the 'C' section is, and dimensioned the e-clip grooves as symmetrical about this centre. I then run a dimension from the center line to the LHS, and from the center line to the RHS.

One of my colleagues has pointed out that this is wrong, but cannot explain how to do it correctly. What would a correctly dimensioned shaft look like?

Thanks.

 

[ctopher]

If you show a centerline, you only need one dim showing the overall dia. It is implied as being on center. If you dim to RHS and LHS, you will have two tolerance buld-up, one side can be shorter than the other, creating almost an elipse. If I understand you correctly. Can you show a pic of your part?

Chris
Systems Analyst, I.S.
SolidWorks/PDMWorks 05
AutoCAD 05
ctopher's home site (updated 06-21-05)
FAQ559-1100
FAQ559-716

 

[ant77]

Thanks for your reply ctopher.

I think my use of the word 'centerline' was a bit misleading. I am not concerned with the diameter of the shaft, but rather its length. The centreline is not the center of the shaft, but the center of the c-section it fits in to. The shaft extends a little more out of one side of the c-section than the other, hence the overall length of the shaft is not symmetrical.

I have created a PDF of the part, showing the way I have typically tried to dimension it (option a) and my latest attempt using GD&T (option b). You can download it at:

http://makeashorterlink.com/?J5C3224AC

(it is only 18kb)

Could you (and anyone else interested) please take a look and comment?

Thanks,

Ant77

 

[Heckler]

Ant77,

This is how I would dimension the shaft. First, the locations of the clip ring grooves need to come from hard datums like in you option B not from an imaginary center of shaft length. I would make Datum -A- one of those shaft diameters. I'm assumming they are all the same size? Do you plan to have an post process like centerless grinding? I would then make your Datum -A- the new Datum -B-. Then the grooves can be referenced back to the center line of the shaft.

Their are a lot of different ways to dimension a shaft. It really depends on overall system design intent.





Best Regards,

Heckler
Sr. Mechanical Engineer
SW2005 SP 5.0 & Pro/E 2001
Dell Precision 370
P4 3.6 GHz, 1GB RAM
XP Pro SP2.0
NVIDIA Quadro FX 1400
o
_`\(,_
(_)/ (_)

"Coming together is a beginning, staying together is progress, and working together is success." - Henry Ford

 

[ant77]

Heckler.

Understand from what you're saying that the primary datum (a) should be the axis of the shaft, and the secondary datum (b) should be the datum I'm currently showing - thanks.

The datum I've attached to the 77 dimension refers to the centerplane of the 77, not the face of the e-clip groove.

My intention is that wherever that 77 ends up on the shaft (controlled by the 10.60 dim) and however big it is, the 93.75 dim should be within +/- 0.05 of it's centreline.

Have I shown this correctly, and is this allowable?

ant77

 

[ant77]

I'd be very keen to get anyone's feedback on the latest version of the drawing. It is posted as a pdf at:

http://makeashorterlink.com/?F575168AC

Thanks and regards,

Anthony

 

[drawoh]

ant77,

It looks mostly okay to me. How are you going to fixture to your datum_B?

I can see what you are trying to do. I would consider using the very end of the shaft as a datum, since this is easy to fixture to, even if it is not particularly functional. A alternative would be to make the far right section of the shaft bigger in diameter, again creating an accessible fixturing surface.

JHG

 

[Heckler]

ant77,

I have a few comments. Datum -A- is referenced to itself in your circularity and straightness callouts. I like to call this a circular reference....a la MS Excel. The E-clip grooves I have never seen the width having the TP callout. It's always the Diameter since you want that to fall right on the centerline of Datum -A-. Do the flats need to be timed/clocked? If so you might want to add Datum -A- to your TP callout? Remember a cylindrical datum feature is always associated with two theoretical planes intersecting at right angles on the datum axis.

Best Regards,

Heckler
Sr. Mechanical Engineer
SW2005 SP 5.0 & Pro/E 2001
Dell Precision 370
P4 3.6 GHz, 1GB RAM
XP Pro SP2.0
NVIDIA Quadro FX 1400
o
_`\(,_
(_)/ (_)

"Coming together is a beginning, staying together is progress, and working together is success." - Henry Ford

 

[ewh]

I agree with Heckler regarding the circular reference. Straightness and roundness two form tolerances which do not need to be related to datums.
As for the true positioning of the grooves, I would prefer parallelism to datum B. I have never seen grooves controlled with true position in this way, either.
In regards to fixturing, I agree with drawoh. Use one end as the datum and control the grooves to it. You may have to add a perpendicularity control of the datum relative to datum A.

 

[drawoh]

Heckler,

My interpretation of the drawing is that Datum_A is only one of five circular surfaces. The others need to be controlled. A similar problem is shown on Figure 5-51 in the ASME Y14.5M-1994 standard, although they did not use the hole as a datum.

The TP callouts look a little weird to me, but check out section 5.10 NONCIRCULAR FEATURES, in the ASME standard. If it were my drawing, there would be dimensions from the datum with tolerances.

JHG

 

[Heckler]

drawoh,

I see your point. I would also use in place of the circularity callout put circular runout since that covers circularity and coaxiality. Or use total runout if you have the budget. I would keep the straightness but remove the datum identifer (not needed with straightness). One more suggestion which totally depends on your design intent but a shaft of this length I would have two datum diameters at either end of the shaft (QA will appreciate that callout). Then make the current Datum -B- the tertiary datum. I would also give the new Datum -B- a runout to Datum -A-. But the big picture is design intent and manufacturing costs.

Best Regards,

Heckler
Sr. Mechanical Engineer
SW2005 SP 5.0 & Pro/E 2001
Dell Precision 370
P4 3.6 GHz, 1GB RAM
XP Pro SP2.0
NVIDIA Quadro FX 1400
o
_`\(,_
(_)/ (_)

"Coming together is a beginning, staying together is progress, and working together is success." - Henry Ford

 

[ant77]

Thanks for all your tips on this shaft.

As always there seems to be a balance between representing your original design intent on the drawing, and describing the part for manufacturing and checking. I have heard of some companies overcoming this by producing 2 drawings for each part, or crowding 'inspection dimensions' onto the original drawing.

This is the first time I have seen TP used for shafts as well, though the standard leaves plenty of scope for it. To play it safe I should probably just use one end as a datum and dimension to it.

Thanks for the tip on 'circular references' - I'll have to do a little more homework here I have read that "...if in doubt, use run out...", which seems applicable here.

Any other thoughts please let me know,

Ant77