Accurate angles from a small part

[kubes]

I have a small collet (a basic cone) for a trim router that I am trying to duplicate. I have an original 1/8" and making a 1/4". It's under .5" x .5, and I need to get the exact angle of the taper. I have tried to measure both ends and have the length to calculate the angle. It is easy to get the large end, but the small end of the taper is my issue. But if my measurements are off a few thou, then the angle will be off. There is no way to get to the inside of the spindle to measure it off the router. I have resorted to trial and error, making a part, dying it, and seeing how it fits. The issue there is it fits to a short flare in the spindle which is maybe .100-.150" on the face. I also asked Porter Cable for the specs, but they were less than helpful.

Of course, the next issue will be setting the correct angle on the lathe....

Any suggestions on the best approach to remaking this part would be greatly appreciated.

 

[Weldingrod1]

I would chuck up a bar, set the taper as good as you can, then blue the router socket and test fit. Adjust the angle to get a good pattern.

Or put sharpie on the OD, put the router on it, give it a twist, and look at the scrape off pattern.

Sent from my SM-G715A using Tapatalk

 

[Dabbler]

You can measure a collet directly if you have one.

Get a pin (gauge pin or drill stem) that is a sliding fit on the collet of choice. Chuck the pin in your 4J chuck, and indicate it in. The slide on your collet and loctite it in place. Sweep along the taper with a tenths indicator ovte a measured length, or, turn your compound so the sweep reads zero along the taper. This will allow you to cut the same taper or close enough it won't matter.

 

[ErichKeane]

I'd hold in a lathe or mill, then use two indicators to measure the change in radius vs the change in length.

 

[extropic]

What @Dabbler said except no need for the Loctite if it fits snuggly on the pin.

You must to have the contact point of the DTI and the lathe cutting tool both PRECISELY at the height of the center of rotation. If your DTI contact point is high or low you will be setting your compound to a greater angle than desired. One way to set that height is to capture a flat piece of light stock between the collet and the DTI contact point. Adjust the height until the piece of stock is precisely vertical. Do the same to both the DTI contact point and the finish cutting tool.

 

[Dabbler]

What @Dabbler said except no need for the Loctite if it fits snuggly on the pin.

I go for a slip fit, so there is minimal distortion in the taper area. I find loctite helps stop it from slipping on the shaft. If it is butted against the chuck jaws, no loctite is needed.

 

[RJSakowski]

Accurately determining the taper angle of a part that small by reverse engineering is a challenging task. I would start by turning a cylinder of known length with a hole slightly larger than the small end of the taper . On the other end, I would counterbore to a diameter slightly smaller than the large end of the taper and deep enough that the collet will seat on the rim of the counterbore.

As suggested by Dabbler, I would fix a close fitting pin in the collet with a glue (I would use cyanoacrylate as the pin would be removable with moderate heat and and the glue can be dissolved with acetone) so the pin is slightly proud of the top of the collet.

To determine the the taper, I would insert the collet into the small end and measure the combined length of the cylinder and seated collet. Then I would flip the cylinder and seat the collet and measure the new combined length. The difference between the two measurements is the distance between the two diameters so you now have the measurements you need to determine the taper.

Concerns are the accuracy of the two diameters and how crisp the edges of the holes are. Any chamfer will offset the seating location. The face of the pin should be flat and perpendicular to central axis of the pin. Another concern is keeping the collet and pin concentric with the two holes. This can be addressed by modifying the cylinder. First drill and ream a hole through the the cylinder for a tight slop fit on the pin to be used in in the measurement and extend the pin past the small end of the collet so it will act as a guide for the seating. Then add a counter bore slightly larger than the small end of the collet. Finally, measurements should be made as accurately as possible. To that end, the combined lengths should be made with a micrometer and multiple measurements should be made and averaged. Counterbore diameters can be bored to fit a pin gage, using care to try to get the same fit on both and the pin gages should be measured with a micrometer to verify their actual diameter.

Done properly, you should now have three values accurate to tenths which is as good as can be done outside a metrology lab. Even so, realize that a stacked error of .0002" on the difference between the diameters and the potential error in the difference between the two combine distances can cause an error of +/- .02º.

 

[homebrewed]

I don't have anything to add to the suggestions regarding the measurement. But your measurement still could be slightly wrong. Once you think you have the right taper angle, you need to test it. Turn a piece with the taper you think is right on the ID, then mark your collet with something like a Sharpie or Dykem. Insert your collet into your test piece and lightly turn it, then take it out and see where the ink was removed. A narrow line at either end of the taper is not good, you want a broader region for a good match.

I never (well, hardly ever) assume my measurements on a piece are correct. Verification is important if it's a relatively high tolerance part. Routers are relatively high RPM tools so you do want a good match.

 

[whitmore]

One other tool in your shop, a drill gage, might be useful: if you fit a mandrel into the collet, then find the
depth the occupied collet sinks into two different drill gage holes, the known diameters (A, B) and the insertion
depth difference C will bear the relation

Tan( 2 * theta ) = ( A - B )/C

with the 'theta' being the half-angle of the turned cone, i.e. what you'd want your
compound set to.

Collets are relatively insensitive to angle variation, one hopes (otherwise they go out-of-compliance
with any variation in the workpiece/tool they hold). You do NOT want tool marks, though,
and a bit of lapping might be appropriate after cutting. Anything that looks like a tool trace is
as unwelcome as a stairstep in a taper.

 

[kubes]

Thanks, everyone, that puts me in the right direction. I have much to learn!