In article ,
Ned Simmons wrote:
In article , dnichols@d-
and-d.com says...
In article ,
Ned Simmons wrote:

[ ... ]

The wheel would have a tapered socket (possibly in a piece
of steel or brass epoxied into the center of the disc) to
match that on the shaft. A nut on the thread would pull the
disc onto the taper. If the angle on the taper is steep
enough (depends on the coeff of friction between shaft and
socket), it will be self releasing or require only a light
tap to release.

That sounds good to me. If fitted to something like that, it
should be very repeatable on any refitting. Of course, the making of
the shafts in reasonable quantities (e.g. onesey-twosey per setup) would
be a bit of a pain -- especially on a small machine. A bit of a runout
gap between the threads and the taper would be a bit easier to handle on
a non-CNC machine, and getting the taper just right might also be a
pain. (It looks short enough to do on the compound, so if you did a
wheel and a shaft with the same compound setup, that should work --
until the next batch which might be different if the compound setting
were to be changed for other work.

Very true, but with a bit of care, the parts should be
interchangeable within a run.

What would be nice would be a reamer to produce the precise
taper in the wheel, so all you would need would be to set the compound
to match each time. Or, with a more gentle taper, you could leave it
set up on a taper attachment. I would suggest using a Morse taper as
the standard, so you could use relatively inexpensive reamers. Probably
a MT-1 (big end) or MT-2 (small end). Any larger would be too big, I
think.

You could have both the taper and thread in the socket,
eliminating the need for the nut, and also supplying the
means to push off the taper, but then it seems the thread
may be more likely to influence the runout.

It also might be hard to start to unscrew, since you've got the
taper fighting you.

Yes, you'd want to be careful about socking it down too
tight.

Bear in mind that the rim of the wheel is bowing strings, which
are vibrating tangent to the rim of the wheel, and the friction is
tailored with a coating of rosin (same as for a fiddle bow), so this is
acting a bit like a tiny impact wrench -- lots of repeated pressure
spikes which would probably tighten the wheel up during playing, and
thus make things difficult to start free afterwards. I would personally
feel happier with a nut -- but one whose gripping surfaces could be
disguised as a decorative feature. For example, start with a cylinder,
and drill eight or twelve holes parallel to the axis around the cylinder
from one end (not going all the way through). Then turn it to about a
20-30 degree taper, until you come out below the holes at the small end.
Then a matching female cone, with two or three inserted pins of drill
rod would give a nice grip on the cone, and it would all *look* like
decoration. The female cone could perhaps be an incomplete circle, like
a 'C', so it could be slipped on over the extension of the shaft to the
final bearing)

Of course, I am not the original poster, I just got mixed up in
this because I've helped build these things in the past (not nearly so
nice as his, though.)

[ ... ]

Surface grinder wheels are often mounted on arbors similar
to this.

Is that why you had the neatly rendered shaft image handy, or
did you do this just for the reply?

No, I did it to save time and avoid ambiguity in my
description. Slow typist, middling to fast CAD modeler
here. It probably took less than 5 minutes to model the
part, paste the screen capture into an image editor, crop
it, and upload it.

O.K. What software are you using? (I'll bet that it is one of
the Windows-based ones, so I can't run it. :-) And how long did it take
you to get that fast?

Enjoy,
DoN.
--
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---