Asymmetric spline used as a datum

[powerhound]

Hi guys,

I need to specify a spline as a datum on a part that we are having built. The spline has a tooth missing so that it can only be put on one way. We also have a spline table on the print that specifies the spline data. How can I call out the spline geometry, not the pitch diameter, as a datum so I can relate the timed features to it? The part is flat and round with the spline in the center. There are features around the outside edge that need to be clocked to the missing tooth. The splines are not square so none of the teeth are features of size.
Thanks,

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[MechNorth]

A graphic would help, but here's my initial shot;
draw a phantom line representing the spline, put a leader to the spline with a note indicating "SPLINE PER DATA TABLE ###", then hang the datum callout from the leader.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[powerhound]

Thanks Jim,

Would then anything that is called out WRT that datum, have to be oriented relative to the missing tooth as shown on the print and geometrically toleranced?

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[MechNorth]

Honestly, I'm just not picturing what you're designing. If you want private feedback wrt an actual drawing, you can e-mail me, and I'd be happy to help. It's jdsykes@ my profile domain, or jim@ the tec-ease acct.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[Belanger]

A datum is a plane, axis, or point. I don't foresee any way for you to identify the spline as a datum feature which includes the clocking aspect of the missing tooth. Instead, the axis of the spline would be a datum, and the opening created by the missing tooth is a separate datum. I'm sure there's more to your issue, but that's my initial take.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[MechNorth]

J-P, a datum feature can be defined by anything that is mathematically definable; the conventional datums are plane/axis/point. '09 even spells out the use of complex datum features.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[Belanger]

But I said that a datum is a plane, axis, or point. Not datum features. (Of course the datum could be a combination of plane/axis/point, as per page 50 of the standard.)

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[MechNorth]

Fair enough, JP. How then do you reconcile a sinusoidal datum feature (as an example) to a single (or combination of) point, axis or plane? Perfectly valid datum feature, which can fully constrain the dof. Does it matter if it's a point/line/plane? Not so much. Unfortunately '09 doesn't provide any guidance for these types of datum features.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[Belanger]

Not sure what you mean. A sinusoidal feature can create a datum plane by sandwiching the top and bottom (peaks and valleys). Or, if the sinusoid wraps around in a circular shape, it would create an axis.
I'm trying to picture a sinusoidal feature that fully constrains all 6 dof. Do you mean a 3D sinusoid, like a topographical image? Or a golf-ball shape?
I'm of the school that if the standard doesn't provide any guidance at all for a particular shape, then datum targets is probably the best way to go.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[MechNorth]

3D topo, like a carton for a flat of eggs. All 6 dof constrained.
Surprised by your last comment; figured you for a bit more adventurous sort. Problem with using datum targets is illustrated in the flatt example. The actual part doesn't have to contact every point on the simulator, but enough to immobilize the part. How can you ensure that the specific targets you select are the ones to make actual contact? Haven't used this much, but definitely was the only way for a couple of clients.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[Belanger]

Sorry if I seem like a stick in the mud. I know you're a fan of "extension of principles," and I can be too, as long as the standard doesn't imply a different view. Don't get me started on using TOP for perp only!

At any rate, I see the idea now. For the egg carton / sinusoid, I think the standard does give us the answer ... see paragraph 4.13. A compound curve or contoured surface can be identified as a datum feature. Then, "its datum feature simulator (derived from the math data) is used in establishing the datum reference frame." In a sense, the whole thing becomes a weird datum target, where we don't identify specific points. A little tricky in practice, I admit, but at least 2009 gives us some guidance that was lacking in '94.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[MechNorth]

Yup. Incremental improvements.

I tend toward the "extension of principles" more than some because I seem to often deal with unconventional/atypical stuff (which is the way I like it). Makes for some interesting coaching sessions when I start down one path and then jump to another once I get more info/insight into the client's system.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[fsincox]

I sure like the direction this took. I, like JP, do not see why this couldn’t be handled as he has proposed earlier. I also realize we have not seen the actual part and one thing I constantly want to stress is: ”we need flexibility” out here. I certainly support the “logical extension of principles” and am not sure why it is even an issue or debate. If we can’t extend the principles past the simple parts in the book, what would be the point of it? The book itself states the examples are purposefully simplistic, where as, real world parts can get very complex.
I believe: “If it functions as a datum, it is a datum”. Whether or not the standard can handle this is the standard’s problem, not the part’s. This is what makes me confident, eventually, the standard will have to be adapted because these are not problems based simply on made up theoretical exercises, but a basic realization of “how a part actually functions in the real world”, If the standard can’t explain it now it is the standard that needs to change to match the real would, or it will never be the tool it should be.
Frank

 

[Belanger]

I agree, Frank. I said that I am fine with extending the principles, "as long as the standard doesn't imply a different view." Many of those extensions have led to clarifications and extra goodies in the standard.


John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[powerhound]

Here's the geometry in question:

Notice the missing gap at the bottom. I know I've been saying a missing tooth but it's really the gap that's missing.

I need to be able to associate the cutouts around the edge to the missing gap in the spline. I want to use the spline as the datum if possible.

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[pmarc]

powerhound,
Is there a chance to see spline tooth/gap geometry more in detail?

 

[MechNorth]

Powerhound,
The spline as a datum feature isn't the easiest for manufacturing or inspection. Easiest would be using the minor diameter as primary and the large tooth as secondary. In that way, the orientation of everything referencing that datum would be controlled relative to the extra tooth (your second question). If you want the orientation to be controlled exclusively by the extra tooth, then this is the way to go (though pitch diameter may be better for spline function than root diameter).

That being said, yes you can still use the spline. If the functionality of the spline doesn't allow contact at the major or minor diameter, you can consider using the faces (or partial faces) of the teeth as datum target areas (as JP suggested). This would make the simulator fabrication a little easier as the major and minor diameters wouldn't be tightly controlled as clearance can be built-in.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[DeanD3W]

Powerhound,
If the tolerances are not too tight, it may be feasible to make a very nice, tooling tolerance version of the splined mating shaft to use as a datum feature simulator. With a very very slight taper and just a bit of force (not too much, to avoid deforming the part you're measuring significantly), you may sufficiently constrain the five degrees of freedom that the entire spline as a datum feature would constrain. The protruding end of this shaft could include a very precise and capable rotation stopping feature that aligns with the gap in the spline pattern and a sufficiently long, precise cylindrical surface to allow finding a useful axis for the spline (could we call this the splines "overall axis"?).

If the production quantity is low enough and the tolerances loose enough you could even consider a simulator with it's splines made of delrin or other compliant material to allow a "pushed together" fit with no rocking between the splined part and simulator. My guess is that the tolerances may be tight enough that a 440 stainless or tooling steel simulator will be needed though, so the level of precision needed to get a good fit may be high. On the small end of this tooling piece you could even try something like wrapping it with a bit of teflon tape (as used in pipe threads) to sufficiently remove rocking between the part and simulator. Another approach might be 440 ss or tooling steel with tapered nylon insterts dovetailed into three slots, 120 degrees apart, cut into the simulator's splined surface. The inserts would then lock out rocking while the splines of the simulator would index the part.

To designate the spline as a datum feature I think a leader with an "all-around circle" added where it bends, and a datum feature label attached to the leader's horizontal leg, would be clear enough. To make sure this is not interpreted as the spline's OD/Major diameter it may work well enough to terminate the leader in a valley between two teeth of the spline, or better yet on a face that creates one side of a tooth. If this is too subtle then a note to describe exactly which degrees of freedom the spline datum feature is to constrain may be worthwhile.

Per Y14.5-2009 I would add a set of X, Y, and Z axes to make it as clear as possible that the gap in the spline is indexing the datum reference frame about the spline's axis. The coordinate system/DRF axes would enable the note mentioned above to include explicit naming of the degrees of freedom constrained.

Dean

http://www.d3w-engineering.com