Coincidence between a hole in a shaft and a tooth on the outside.

[James7443]

I have a shaft that has five teeth equally spaced around the outside of the shaft. There is also a hole drilled thru the shaft perpendicular to the axis. The hole is used to index the shaft in reference to one of the teeth. The origin of this part dates back to the early 60's in which the print then had a note stating "centerline of the tooth thru the axis of the shaft and the centerline of the hole to coincide within +/- 0 degrees 15 minutes".

I am trying to determine the correct way to define this using GD&T. Attached is a sketch of my first thought, but I am not sure if it is correct or the best way. Datum A is the center of shaft bore. Datum B is the end of the shaft. Datum C is the profile of one tooth which should give me the axis of the tooth. The .002 parallel to C in the control frame is equal to 30 minutes of angle. Which now I am thinking should be .001 or 15 minutes.

Does this sketch seem correct or is there a better way? Or should I put a note under the control frame stating axis of C and axis of hole to be coincident within +/- 15 minutes.

Thanks in advance.
Ron

 

[CheckerHater]

You will have to emphasize that you are using axis of the tooth in your GD&T scheme.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[James7443]

Thanks. I can add that dimension. The sketch isn't exactly correct. The tooth doesn't have any straight edges. More of a light bulb shape which is why I was going to attach the datum to the profile edge.

I actually was just thinking about changing the control frame to three lines.

True position| .002| A | C | This line to prevent the location from getting too far from axis of tooth or bore.
True position| .005 | B | This line to give a little more freedom along the length of the shaft.
Parallel |.001 | C This line to control angle to the tooth.

 

[Tarator]

I think this should be the correct way.... without datum C, you can control concentricity/coaxiality

 

[3DDave]

I think the second segment should be [A|C] because [C] by itself has no rotation control - picture the number of orientations one might close a caliper on to measure the width. Ordinary gear measurements, not covered by Y14.5, are made in reference to designated features for this reason. Only if the teeth are rectangular is the orientation of the tooth width unambiguous.

 

[tbuelna]

Not clear how the CL of the tooth profile is established. Since it appears that the opposing tooth flanks have some form of convex profile, using a parallel pair of lines symmetric about datum A and tangent to the tooth flank profiles does not ensure the tooth profile CL actually passes through datum A. Instead I would suggest using the .094" radial hole as datum C, and applying a profile tolerance to the 5 tooth flanks based on datums A,B,C.

 

[CheckerHater]

picture the number of orientations one might close a caliper on to measure the width

Actually, not so many


"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[3DDave]

CH - that's not the zone that the previous depictions of [C] callout represent. I believe you are pulling my leg.

Even a tooth thickness caliper doesn't have a unique orientation.

 

[CheckerHater]

All I am trying to say is that industries dealing with gears have all necessary tooling to measure exotic parameters like runout of pitch diameter or chordal tooth thickness.
I assume you have no problem dealing with concept of hole axis, which is imaginary, just like CL of tooth profile?

Naturally, all the sophisticated instruments leave the place for simple techniques like this:

http://www.eng-tips.com/viewthread.cfm?qid=111643

Let me repeat, dealing with tooth thickness / center plane may be easier than you make it to look like.


"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[3DDave]

Chordal tooth thickness can be sufficiently determined even if the orientation of the caliper is off by a few degrees. Typical measures of gear tooth profile are done relative to a defined axis, which is why I suggested [A|C] instead of [C] alone. And that is what the mechanisms the link you provided depend on - a defined axis in conjunction with a locating element.