Making Large Gears on an Atlas Horizontal Milling Machine
George Langford
georgesbasement.com
Method 1 - January 21, 2005
Index a large gear's teeth   The Atlas Press Co. 10 inch lathe that I purchased for my son had a form of metallurgical cancer in the Zamak aluminum-zinc alloy gears.  Several teeth fell right out of the main bull gear on the headstock, and closer examination revealed swelling and cracks everywhere on the gear.  So I had to machine a new one to make the lathe functional.  Even though Atlas's owner, Clausing, is extremely cooperative and helpful, a new wrought aluminum gear from them was way outside my budget.
   Problem was, the gear was so large that I could not swing it on the Atlas index head and still get a gear cutter over the top.  So the gear had to be machined horizontally on the Atlas's meager rotary table, which has no indexing capability whatsoever.  My solution is shown at left.  The bad gear is mounted flat on the rotary table.  Above it is the new gear, with about three-quarters of its teeth already cut.  In the foreground is the Atlas lathe's second back gear, removed from the gear cluster for the time being.  The index pawl is at the far right.  The pawl engages the teeth of the intact second back gear, and that gear's teeth average out the errors and bridge over the gaps in the defunct bull gear. The gear was about three-quarters finished when I took the picture.
Old and new gear teeth   The new gear teeth are shown above the bad bull gear's teeth at left.  I offhand ground the fly cutter, so these are not the quietest teeth on the planet.  But they work great.  After a little running in, the meshing quieted down.  I did take a lot of trouble to make the fit of the bore of the gear blank very close to the seat on the outside of the spindle and to mount the blank concentrically on the rotary table, and that worked out well.
   The idler gear rotates on a shaft mounted on the I-beam; I left it free to rotate, held only by the index pawl.  Of course, once I had indexed to each new tooth, I took up the slack and then tightened the clamps that kept the rotary table and the gears bolted to it from rotating. So everything was nice and tight when I started each cut.  There's lots of impacting from that fly cutter, seen at top left of the image at left.  Note that I fed the knee upwards to get a smooth motion without the force of the cutter tending to lift the knee or saddle.
   The principal limitation of this technique is the reach of the fly cutter.  It would be more rigid to use the Atlas's overarm to support the outboard end of the flycutter.  One might make a gear a foot across; I haven't checked what the exact limit might be.
Method 2 - June 13, 2013
All set up and ready to start   Method 1 was necessitated by that cancerous bull gear. The present Method 2 came out of my decision to complete the set of metric change gears that I started about 25 years ago. I started with the composite 127/100 gear set from South Bend lathe that can be seen at the bottom of the stack in the image at left. The large aluminum plate near the top of the stack is the intended lead screw gear, which is to be a copy of the 100-tooth gear in the 127/100 composite, but with a 0.5625 (i.e., 9/16) inch bore and 0.125 (i.e., 1/8) inch keyway. You may also notice that I've graduated to a real gearcutter, which in turn necessitated making the 0.875 (i.e., 7/8) inch arbor seen at left.
Flycutter setup   Another task was making the four smaller gears (18, 20, 22 & 26 teeth) that are also needed to cover the expected range of metric thread pitches. I made the first of these with the same flycutter (albeit ground to match the smaller gears in the lathe's normal English change gears, which of course are all interchangeable with the same 18 diametral pitch, 14-1/2 degree pressure angle as the metric change gears). Shown at left is the Atlas indexing head and the master gear (twice as many teeth, hence skipping every other tooth with the index pin). The others I made with gearcutters.

Roughing out the wrought iron arbor   To make the 7/8 inch arbor, I hunted around my basement and found a likely looking piece of ferrous material, but on the first pass it became clear that I was machining a piece of wrought iron. Except for needing to sharpen the high speed steel cutting tool frequently, it was easy to work accurately; note the pile of short, curly chips. Not shown is the splinter of slag that I got while caressing my fine work ...
   The images at right show the chip pile in greater detail as well as my process for making the 5/8 inch clamping nut for the arbor.
Closeup of the pile of wrought iron chips
Arbor clamping nut setup
Setting the angle of the six-sided nut
   The series of images below show the completed vertical arbor, which consists of two main pieces: A base, made from a piece of 1x3x8 inch cold rolled steel, which only needed a receptacle bored out to a close fit to the lower clamping nut of the gear carrier, a keyway to fit the Atlas main table, holes for the clamping screws and tee nuts that hold the base down, and an adjustable index pin to accommodate the master gear. The adjustable index pin is held by a sliding member that is keyed to the tubular support arm.  When I cut the present 100-tooth gear, I moved one of the spacer gears underneath the blank, the better to support it against the downward force exerted by the gear cutter. Making non-cluster gears will necessitate making a new gear carrier to fit the 9/16 inch bore of separate change gears.
The complete vertical arbor
Vertical arbor sections shown separate
Vertical arbor with gears in place
Bottom view of vertical arbor
Cutting the Nth tooth   Here's the setup, shown towards the end of the three-hour tooth-forming process. The two strap clamps are keeping the gear carrier and master gear from rotating and take the load off the index pin. There was some vibration of the gear carrier because of deflection of the 100-tooth master gear (second from the bottom) whose web is rather thin. Next time I will wedge spacers between the two gears or separate them entirely and use a flat-plate spacer on top of the master gear. The cutter passed through the first tooth gap cleanly after the gear was finished, proving that nothing had slipped.
   In practice the 127 tooth member of the 127/100 compound gear is driven by the lathe's stud gear (which rotates at the same rate as the lathe spindle), and the 100 tooth member of the 127/100 compound gear in turn drives the 0.125 inch pitch lead screw, either through a 72 tooth idler for the 6.00 mm through 2.00 mm thread pitches, directly for the 1.75 mm through 0.40 mm pitches, or with a 4:1 reduction by the 72/18 tooth compound gear that is part of the standard English thread change gear set. The 100 tooth change gear that I just made is used alternately with the 80 tooth change gear from the English thread change gear set on the English lead screw.
   I did not make a proper metric change gear setup for the image at right, which was intended simply to demonstrate to myself that the 100 tooth gear meshes correctly with the other 18 DP, 14-1/2 PA gears without excessive runout. OK !!
100-tooth gear in place on the feed screw
Note: Most of the above images can be viewed in greater detail by right-clicking with the mouse and selecting "view image."