---- David Beierl <dbeierl@ATTGLOBAL.NET> wrote:
 For
> larger ones solid-handled affairs resembling die stocks, with a
> two-jaw chuck operated by screwing in one of the handles.

That is my tap handle.

> >   Working very slowly, with oil, I was able to keep the tap moving
> > without it going off course.
>

> This is purely a matter of dimensions.  The socket is a solid
> casting; there's no point where the soft metal stops and hard takes
> over.  If you have a small enough hole left to marry to the splines
> without them sticking out too far or the arm bottoming on something,
> it will work.
>
> Were you certain that the splines were shiny-clean?  The old splines
> from the arm shear off and stay in the grooves between the splines on
> the shaft, and must be removed.

I recheck and re-recheck.  I washed with oil and a wire brush.  but I'll redo that, and check the thing out with my botanists lens.  then try to marry the two parts again, once I know it is clean.  Nothing to lose by doing it, that's for sure.
>
> >BTW, Alistair, I did use a prick punch and then a larger center
> >punch.  The stud (which was cut down actually into the conical shaft
> >in order to get a flat surface to work with) showed no tendency to
> >move, though I whacked pretty good, but with a light ball peen hammer.
>
> Light hammer is the key - you want to upset the surface of the shaft,
> not move it.  In electronics terms you're matching the impedance of
> the driver to the load.
>
>
> >So far as center is concerned, I hit what I perceived as center with
> >the prick punch, and examined the stud with a 3x botanist's field
> >lens.  I accepted that I had hit center,
>
> For a mounting bolt that has to be plenty good.
>
>
> >Though I did not accomplish this job as I'd hoped, I really do
> >appreciate all the pointers and advice given.
>
> Re-check those male splines before you give up.  There is a
> deterministic reason why you didn't get splines cut on the female
> part, and it's not that you ran out of cuttable metal.  It has
> something to do with dimensions/interference or insufficient clamping pressure.
>
> Yours,
> David

--
David McNeely