Approaches to turning a spool-shaped object (capstan drum)?

[Huub Buis]

Thanks Hubb--I should have mentioned I'm limited to a manual lathe, but it's interesting to hear how you would approach this part with CNC.
-Mike

On a manual lathe, I would split the part in 3, the 2 flanges and the mid portion. I would connect them using a press fit or a sloppy fit and glue them together using epoxy, or soft soldering using 3% silver solder (available in a general hardware store), or screw them together using a small pitch thread.

I am puzzled about the design,
The big key suggest it has to transmit a lot of torque but a smooth surface can't transfer a lot of torque. May be you could "lock the part"on the shaft by just pressing 2 o-rings together in the inner (large) boring.

 

[jwmelvin]

-would a tapered, pressed-on mandrel actually firmly (and concentrically) hold this part? The typical taper on the mandrels I see is 0.0005/inch.

I’d use a straight mandrel, with tailstock pressure. You could use a stepped mandrel with the 1” bore on the tailstock side. Could also put the keyway in after boring and use a key in the mandrel.

 

[rabler]

would a tapered, pressed-on mandrel actually firmly (and concentrically) hold this part? The typical taper on the mandrels I see is 0.0005/inch.

I've never used a tapered or pre-made mandrel. Just spot make them as needed. For this part, I'd start with a piece of 1.5" round bar in a three jaw chuck. Turn a 1.001" diameter stub so the part sits up against the 1.5 to 1" shoulder. Use emory paper to bring the mandrel to a tight but not press fit. Then either male thread the 1" end for a nut, or drill and tap it for a bolt, along with a washer to pull the part into that shoulder. Preferable to do all of this in one setup without ever removing the mandrel from the 3-jaw so it is dead minimal runout.

If the fit is a bit too loose, a little cyanoacrylic (super glue) to hold it in place supplements the nut/bolt, and a little heat from a propane torch to release it.

Either a nut or bolt works to hold it on, but if the length to diameter ratio is too large the mandrel itself can flex. In that case I prefer to have a male threaded mandrel where I can center drill the end for a tail stock. I suppose you could center drill the head of a softer bolt too, never tried that.

 

[MikeMT]

Thanks a lot for all the ideas everyone--very helpful, I appreciate it. I feel like I've got multiple good approaches to consider now and this is very educational.

----
rabler, the idea of center drilling a male threaded mandrel end is very clever, I like it!

The big key suggest it has to transmit a lot of torque

Well it's actually just because that's the size shaft I've got available to put this on--hobby project

I would split the part in 3, the 2 flanges and the mid portion. I would connect them using a press fit or a sloppy fit..

I am intrigued in the split the part up approach as turning the tapered section between two shoulders seems to have some potential challenges (at my skill level anyway)--although I recognize the split up approach could create its own challenges/complexities. When you say press fit, what is being pressed into what? Just the three loose pieces onto the shaft, or would there be additional bosses or pins to mechanically align/hold the parts together?

 

[mickri]

You have not told us what the capstan is to be used for and what it is attached to. Also how it is attached. My guess is that the dimensions of the large hole are not critical. It only needs to be big enough for a socket to tighten whatever holds the capstan on. Also doesn't have to be absolutely concentric to the 1" bore. Also the outer portion of the capstan probably doesn't have to be absolutely concentric with the 1" hole either. If some type of rope/cable is going to be wrapped around the capstan then it might be good to knurl the outer portion. Based upon my guess the only truly critical dimension is the 1" bore and its keyway. Lots of wiggle room in all of the other dimensions.

This is how I would make the capstan. I would start off with the raw aluminum being about 1.5" longer than the finished part to hold it in a 3 jaw or a 4 jaw chuck. I don't think that it matters what you use. First bore the 2" hole. Then swap the work end for end, face off the end and bore the 1" hole. Make the keyway. Mr. Pete has a video on how to make a keyway on a lathe. Then turn the outside of the capstan to shape. Part it off. Clean up the outer face. You are done.

 

[MikeMT]

Thanks for the thoughts mickri--

You have not told us what the capstan is to be used for and what it is attached to. Also how it is attached.

It would be used for pulling on a rope, and attached to the shaft of a right angle gear reducer, secured with a set screw.

Based upon my guess the only truly critical dimension is the 1" bore and its keyway. Lots of wiggle room in all of the other dimensions.

Good guessing. I think you are absolutely correct here that there is lots of wiggle room. Though I am certainly interested in the full spectrum of approaches to turning it, from highly accurate to less so.

This is how I would make the capstan.

Appreciate the input, I like this approach.

 

[mickri]

If you really want everything to be as close to concentric as possible the first step would be to turn the outside between centers. Then dial in the work in a 4 jaw with a steady rest on the tail stock end. Bore the 2" hole. Swap end for end and bore the 1" hole. Broach the keyway. Replace the steady rest with a bullnose live center. Then turn the outside to shape and part off.

What will all this extra work get you? That will all depend on how close you can dial in the work in the 4 jaw and steady rest. Probably cut the runout to a couple of thousands verses 5 to 6 thousands.

 

[Huub Buis]

I am intrigued in the split the part up approach as turning the tapered section between two shoulders seems to have some potential challenges (at my skill level anyway)--although I recognize the split up approach could create its own challenges/complexities. When you say press fit, what is being pressed into what? Just the three loose pieces onto the shaft, or would there be additional bosses or pins to mechanically align/hold the parts together?

The 2 flanges would get a "male boss" and the inner part a "female boss" on both sides. Because the parts are symmetric, alignment pins are not needed. After the parts are pressed together, I would press the part to the chuck using the tail stock and a live center. Than use some sand paper to eliminate the seems and finally polish the whole thing.

Have you thought about 3D printing this part (prototyping)?

 

[Larry$]

My only experience with capstans was in the Navy. They were big and turned relatively slowly. The line (rope) surfaces were smooth and uniformly shaped so the line would always travel toward the center whether it was taking up or letting out. The surfaces need to be smooth so they do not abrade the line. The line is continuously slipping as the crew takes up the slack to tension and pull the line in or slacks off and allows the line to go out. Several loups of line are wound around the capstan to provide the correct amount of control. The canstan turned at a uniform, constan, speed. In addition to docking the capstans were used for highlining loads from ship to ship while at sea.

 

[MikeMT]

The 2 flanges would get a "male boss" and the inner part a "female boss" on both sides…

Got it, makes good sense, thanks for the explanation Huub. A lot of good strategies to think about now.

My only experience with capstans was in the Navy…

Cool Larry$, thanks for the insights—now those were no doubt some “real” capstans! I’ve seen a lot of variations on the basic concept, seemingly tailored to slightly different use cases. If I had the time I’d love to try making and testing a few different versions… As Huub suggested maybe that’s a good job for the 3D printer…