Recently, I noticed the drawbar on my Millrite vertical mill was beginning to feel soft when I tightened it. This MVI has a R8 spindle.
The drawbar had been repaired by the previous owner, the repair consisting of cutting the 7/16-20 threaded part off, drilling and tapping the bar, and loctiting a piece of 7/16-20 allthread in place of the missing part. This worked for many years, but recently the feel began to change, so I figured it would soon break.
DC Morrision does carry the drawbars, for $145, which seemed a bit high for something so simple, so I decided to make one. This had to be my immediate project, so I could finish before the incumbent drawbar broke, as I would prefer to mill the flats for the 5/8" square where the crank handle goes, not to file them by hand, slowly.
The the standard drawbars are made of ordinary mild steel, and these were breaking, so I wanted something a bit stronger, so I would never have to do this again.
I settled on 1144 "stressproof" steel, which is easily machined and is about twice the strength of 1018 steel. Given that 1018 worked for many years, I bet 1144 will never fail in this application.
The bar is made of 0.500" diameter rod, with a big machined nut on top. The 6' round bar from McMaster rang when struck, unlike 1018, which makes more of a clunk sound. I assume that 1144 is stiffer than 1018.
Threading on the Clausing 5914 was done at low speed using the back gear, with brushed-on black sulfur oil lubricant. (I didn't trust my usual oil emulsion coolant for such a severe machining operation. And the very low speed places less stress on the machinist as well.) Because the 7/16-20 thread section (that engages the collet or whatever) is at least three rod diameters long, I center-drilled the end and stabilized the rod with a live center. Even so, the threads were larger away from the chuck due to deflection of the rod. I fixed this up by running a thread-cutting die over the threads to clean them up. The tips of the threads remained slightly feathered, but some filing left nice smooth threads.
The glorified nut that will be at the top was also made of 1144. The
1144 machined very easily, as advertised. When I was whittling the 1.5" diameter rod down to rough size, taking 0.250" of diameter off per pass under power feed, I got very even coil-spring chips 1 to 3 feet long, that turned straw then dark blue color as they flowed away from the cutter. After cooling, the chips were very strong and springy.Tapping was done by hand using the unpowered lathe as a tapping fixture to hold the work and guide the tap straight into the drilled hole to be threaded. A bottoming tap was used to complete the thread. Tapping went without difficulty, although as might be expected the tapping force is far greater than for 1018 steel.
The union between the nut and the bar is an inch of 1/2-20 thread. The main forces on it are the pull when a R8 tool is secured in the spindle, and the hammering on the nut to release the tool. The force directions are opposite, which can be a problem for Loctite, so I arranged the shape of the rod tip to rest solidly on the bottom of the threaded hole, and not to run out of thread too soon. Then, metal-to-metal contact between threads carries the securing force, while metal-to-metal contact between rod tip and hole bottom carries the hammering forces, so all the Loctite 271 has to do is to prevent unscrewing.
For the record, the original design uses a #3 taper pin to fix nut to rod, and this had become loose over the years, so I decided to use Loctite. If this loosens (unlikely), I can always install a taper pin.
Assembly. There are two objectives. First, ensure that the entire joint is filled with Loctite 271 (the original red stuff). Second, ensure that the rod is forcibly bottomed in the hole, so that later hammering won't crush the Loctite. The rod and nut were both washed with acetone and blown clean, to remove the cutting oil, the rod thread was liberally coated with Loctite, and the nut was run on and off twice to distribute the Loctite evenly. Then, the rod was clamped in a vise and the nut tightened with a wrench. Excess Loctite was wiped off.
Loctite 271 takes 24 hours for a full cure. To ensure the full thread-to-thread contact, curing is being done with the assembled drawbar installed in the mill, in operational configuration, securing a big toolholder to the spindle.
Joe Gwinn