Another Ball Turning thread - just what we needed

[kvom]

Having read all of the ball turning threads of late, I decided to have a go using the "human CNC techique" this evening. I ran Marv's program to produce cutting schedules for both .04" AND .02" increments for a 1" diameter ball. I ended up using the .04" schedule.

My camera is in for repairs, so no photos currently available.

The ball is intended to replace the plastic on the end of the handle of the QCTP, which has a 3/8-16 thread. I found a piece of 1" dia. 6061 aluminum, chucked in a 5C collet, and turned down a 1.5" length to 1/2" diameter. This was drilled and tapped. I then reversed the piece chucking the handle in a .5" collet, and faced the exposed giving a 1" length of 1" diameter which would become the ball.

I followed Marv's schedule using a LH bit for half the ball, then mounted a RH bit to form the inner half. Perhaps using a parting tool would be quicker. The cutting went very fast, less than 10 minutes, as aluminum cuts easily. Subsequently I measured to find that my diameter was .997 and my length was 1.02, so I was .005" oblong.

Now I coated the ball with Dychem, and started smoothing with a file until none of the color remained and I could not see any lines from the "slices". I followed this up with 3 grades of emery cloth, 2 grades of Scotchbrite, and finally Mother's metal polish. Total time from first cut to polish was about 45 minutes, so probably equal to setup and cutting time with a ball cutter.

However, my result was not exactly spherical, with the end being rather flat. I suspect that to get a really good sphere a lot more filing would be required, with reference to a radius gauge perhaps. For the purpose of a handle (which I suspect a great many home shop balls are used for), my "ball" is perfectly OK.

Note that in running Marv's program, I selected the "x-axis" method. The alternative is "angle", which may give more cuts on the ends, where it's more critical to get a proper shape.

 

[mklotz]

Note that in running Marv's program, I selected the "x-axis" method. The alternative is "angle", which may give more cuts on the ends, where it's more critical to get a proper shape.

That's exactly why the angular method is included. If you have problems concentrating, the equal-x increment method will help but the equal-angle increment method produces better results, especially at the top of the ball.

I'm pleased that you tried the incremental method. While I built a ball turner, I almost never use it and, in retrospect, if I knew then what I know now, I would never have built it.

Almost all ball shapes made by the hobbyist are decorative, not precision, so the very minor inaccuracies that may result from the incremental approach are pretty much moot. Moreover, just how many balls are you going to make in your career? One can make a lot of balls incrementally in the time it takes to build, fine tune, setup and use a ball turner.

I'll admit that certain concave surfaces are better done with a turner (although many can be done incrementally) but, as my PROFILE program demonstrates, a vast variety of shapes can be roughed out incrementally in a relatively short time. Not everything we do requires a new, complex, designed-to-purpose tool that invites deviation from the project at hand to construct and debug.

 

[Maryak]

For wood or plastic balls on the end of handles just use a well sharpened Wad Punch or piece of steel pipe shaped and sharpened on the end like a Wad punch. Use a rocking round motion of the punch whilst the lathe is revolving on a roughed out by hand ball to give the finished ball.

Quick and dirty but it works.

Best Regards
Bob

 

[artrans]

this thread sucks with pictures ;D just fiqured I would beat brass to it although he has that cool bill board which I can not copy to here for some reason anyway just jokeing

 

[Brass_Machine]

This one?

 

[itowbig]

how do you get those . if you please

 

[Maryak]

Try this,

http://madmodder.net//Smileys/default/worthless.gif

Now back to the topic

Best Regards
Bob