Wheel Cuting thin tooth?

[Phil PParticipant @philp]

In the top and bottom scans, something looks a bit odd about the tooth profile at around the 1.00 o’clock position to me.

 

Or is it just distortion in the scan ?.

 

Phil

 

[johnp10Participant @johnp10]

Hello Mr Hall,

I was interested to read that you are building my skeleton clock and would be interested to know how you are getting on with it

May I invite you to contact me on johnp10@virginmedia.com

I would like to hear you experience of this project .

Regards.

John Parslow .

[Sub MandrelParticipant @submandrel]

I have been thinking. the holes in teh wheel are OK, and its direct dividing.

 

Somewhere I think it was mentioned that the last tooth was wrong. Perhaps it was really the first?

What if the first hole is selected in a different way to the second, third etc., meaning there is some backlash that disappears – perhaps over a series of cuts?

 

Neil

[RobCParticipant @robc77385]

Peter,

 

One thing which has not been mentioned so far (I believe) is how you are achieving your depth of cut on each tooth.

Are you cutting to full depth in one pass, or are you taking multiple passes at different depths?

If you are taking a single pass, then we are all on track with the challenges and can safely ignore the rest of this post. But if you are taking multiple passes, there is a possibility of errors creeping in depending on your technique.

I have only ever cut two gears myself, one of which is sitting on the table beside me (I am building Peter Heimann’s regulator) but I have noticed a few things with gear cutting.

If you are taking multiple cuts, you should cut a full 360degrees of teeth before altering the depth of cut for the second (or subsequent) passes. If you don’t there is a possibility of different width teeth due to positional issues. (Not explained well, but you probably get the idea).

As to the others who mentioned the number of teeth requiring a correct sized blank; the blank size only determines the finished tooth width, which is obviously important from a meshing perspective. A correctly working indexing system will produce the designed number of cuts for any size of blank, but the teeth will only be the correct form for the appropriate blank size. Too large a blank and the teeth will be too wide, too small a blank and the cutting process will make them too short and thin.

Rob

 

[Sub MandrelParticipant @submandrel]

On reflection, I think Blowlamp has the answer.

 

If the lathe creeps a tiny bit during each cut, then the final tooth will be slightly too far round, or not far enough, and teh error will be distributed more or less evenly across all teeth.

 

PLEASE take his advice and fit the detent to the bed of the lathe and try again.

 

Neil

[Michael GilliganParticipant @michaelgilligan61133]

This is an interesting hypothesis; but it does beg the question as to how Peter subsequently managed to cut good wheels with the same set-up.

One simple test for Peter [in the absence of a ShadowGraph projector]
Try holding the cutter against several of the cut spaces in your bad wheel.
… if it is a snug fit, then we still have a mystery.
… if it is loose, then creep may be the answer.

One thought:
I note that the indexing arm is rather long and slender …
Assuming that [to remove backlash] you are pre-loading with the typical “weighted string around the chuck” … do make sure that you are putting that indexing arm in Tension, not Compression.

MichaelG.

[Author]

Posts