The purpose of this document is to record the removal and replacement of the spindle of a US Burke Millrite model MVI vertical mill, which was built in 1965. This saga records my efforts in August 2010, triggered by increasingly disturbing noises coming from the spindle. In this saga, it is assumed that the reader has a copy of "Millrite Maintenance Instructions and Parts List", 56 pages, available gratis from The D.C. Morrison Company, 201 Johnson St., PO Box 586, Covington, KY 41011, USA, (859) 581-7511 voice, (859) 581-9642 fax. (They are open 8-5 ET, M-F.) Item and page numbers herein refer to this manual. 1. Remove the cone pulley assembly (page 16, and item 15 on page 38) on the top of the spindle: First, remove the drawbar. The manual suggests (on page 17 bottom) that one should completely remove the motor, but this is awkward as the motor is heavy and my shop is crowded and the ceiling is low, so I instead moved the motor out of the way. The motor assembly is held to the head plate (item 1 on page 38) by means of two clamp studs, one passing through a round hole and the other through a long curved slot, both in the head plate. Items 19-22 on page 38 show by example the arrangement for both studs. Removed the stud that passes through the long slot by removing the setscrew and unscrewing the stud. Loosened other stud. This allowed the motor assembly to be swung out of the way without removal. But be careful not to swing the motor too far, or the weight of the motor amplified by the newfound mechanical leverage may break the head plate or motor mount castings, which are aluminum alloy. One may unscrew the threaded stud by tightening two 5/8-11 nuts against one another on the stud, and then turning the stud using a wrench on the inner nut only, so that removal torque tends to tighten the nuts against one another. Stud installation is the reverse, and one puts the wrench on the outer nut. While the stud was out, I filed a flat on the threads where the setscrew rested, so the deformed threads would not further cut the aluminum of the motor mount. The cone pulley is held in place by a ring nut (item 16 on page 30). The ring nut has an outside diameter of 1.250", a setscrew and a 0.182" diameter radial hole that accepts a radial-pin spanner wrench. (I used an Armstrong model 34-351, which has a smallish pin for the hole, but it worked without difficulty. Bought from McMaster-Carr.) Spindle rotation is prevented using the pulley lock (item 7 on page 38). Don't lose the little brass disk or plug under the setscrew, used to protect the spindle threads, but not shown in the drawings. It's best to loosen the top ring nut on the spindle now, before removal of the cone pulley, so the cone pulley can be used to prevent spindle rotation. This ring nut is not shown clearly on any drawing, but must be removed to permit removal of the spindle from the quill. The best depiction is on page 14, where it is shown as a thin hex nut at the top, upon which the drawbar nut (item 105 on page 30) rests. Because of drawbar tightening, this nut may over time become very tight, too tight to be removed by wrench alone without damaging nut and/or pin wrench. My solution was to use an air-acetylene plumbers torch to rapidly heat the ring nut (but carefully keeping the flame away from the spindle) until the grease began to smoke (about 450 F), and then quickly unscrew the now loosened nut with big pipe wrench. There were already pipe-wrench marks on the ring nut, and one can see why. Although suspected, no loctite was found. Actually, the pin spanner would have worked on the smoking hot ring nut. The heating must be quite rapid, so the ring expands faster than the spindle, becoming looser in the process. Once loose, the nut should be unscrewed immediately, before temperatures can equalize. Now that the two ring nuts are loose, remove both, followed by the cone pulley and its square key (items 15 and 16 on page 38). Millrite_Spindle_Saga_Photo1.jpg shows the head assembly just after removal of the cone pulley. 2. Removal of the bearing cage cover (item 20 on page 30): Mark these parts for orientation before removal using a scriber or prick punch. Unscrew the three screws (item 21 shown above item 20) and the cover plate (item 20). The whole assembly can now be removed, slid off the spindle spline, at least in theory. In my MVI, the spindle spline had been deformed near the top by use of a pipe wrench, probably in an effort to remove the topmost ring nut without the cone pulley to prevent spindle rotation. After cleaning the spline ribs up with a small file, the assembly (items 13-19 on page 30) slid right off and could be removed from the machine. Millrite_Spindle_Saga_Photo2.jpg shows the head assembly with the pulley sleeve (item 15 on page 30) and associated hardware removed, leaving the spindle spline unencumbered. 3. The quill nut, which is about 3.25" in diameter and forms the bottom of the quill, has a left-hand thread and requires use a face pin spanner wrench to loosen and tighten. (I used an Armstrong model 34-154 for this.) It turned out that the quill nut was loose, which is a problem, as will be discussed later. According to page 17 bottom, the spindle assembly (including the taper roller bearings) is supposed to drop right out with a sharp rap to the top of the spindle (put the top ring nut back before rapping the spindle, to protect the threads) once the snap ring (item 14 on page 30; also shown on page 14 between text items 5 and 6) and pipe plug (item A on page 14 and item 12 on page 30) and the quill nut (item 11 on page 30) are all removed. (Don't forget the ring nut at the top of the spindle, and a wood block below to catch the falling spindle assembly.) This approach ultimately worked, but required an intimidating amount of pounding with a big lead hammer. It would probably have been easier and gentler to use a piece of 1/2-13 allthread rod passing through the spindle bore to a T-Nut in a table slot with a thick washer and nut at the top. Clamp quill and lower table to pull the spindle assembly straight out. Or, clamp table and retract quill. With pressure applied, gentle tapping may be required to get things started. Millrite_Spindle_Saga_Photo3.jpg shows the spindle assembly, complete with the two taper roller bearings, the spacer, and the pretension nut. The thin conical skirt at the spindle nose (to the left in the picture) is the grease slinger. 4. The intent is to remove the two Timken taper roller bearings for cleaning and inspection, and possible replacement. First, the bearing retainer/ pretension nut (item 8 on page 30) must be removed. It has a setscrew (item 10) plus a brass plug (item 9) to protect the spindle threads, and three 0.185" diameter holes to accept a radial pin spanner wrench. Again, the Armstrong model 34-351 radial pin spanner was used. The bearings were supposed to be a "tapping" fit, not a real press fit, so the bearing preload can be adjusted by turning the retainer nut (item 8). However, the bearings could not be removed by any amount of tapping with a big hammer and a brass rod, and had to be pulled or pressed off. Not having a suitable puller or hydraulic press, I had to improvise. What I used was a piece of 1/2-13 allthread passed through the spindle bore with heavy washers and nuts at the two ends. A pusher made from cheap 1-1/4" black iron pipe and cast iron pipe fittings allowed me to apply force to the bearing inner race without crushing the grease slinger (item 5 on page 30). The motion was intermittant, as the nuts were tightened the allthread stretched until sufficient force had been generated to move the bearings, the bearings moved then stopped, so motion was abrupt and noisy. But it worked, probably because this was only a light press fit. When machining black cast iron pipe fittings, the easiest way to hold the workpiece in a lathe is to screw the fitting tightly onto a short bit of threaded black iron pipe that is held in a three-jaw chuck. Millrite_Spindle_Saga_Photo4.jpg shows a cast iron pipe fitting in the lathe, just after machining a pusher end. A number of different pushers were made; this one is representative. The spindle bearings are a pair of Timken taper roller bearings nose to nose kept apart by a tubular mild steel spacer between the two outer races. Each bearing consists of a model 19150 (the "cone", being the inner race, the rollers, and the roller cage) plus a model 19268 (the "cup" or outer race). The bearings used here are Timken Class 3 precision, which Timken recommends for use in machine tools. 5. The grease in the bearings had turned into a gritty black dry cake. After much cleaning with acetone and VMP Naptha and elbow grease, got the bearings clean. The only damage found was a small (about 0.25" diameter) spot of superficial corrosion due to entry of water-based metalworking fluids. If the grease had not degraded, the water would have caused no trouble. A set of new bearings costs about US $400 and the damage was only cosmetic, so I decided to reuse the existing bearings. The spindle shaft had lengthwise furrows from grit trapped between bearing race and shaft scratching the shaft as the bearings were pushed off. It was not possible to remove the accumulated grit until the bearings were removed. Cleaned shaft up with ultra fine sandpaper backed with a sheet of stiff orange rubber gasket material, the stiff gasket material ensuring that the peaks of the furrows are selectively worn down to the original surface, followed by polishing with crocus cloth. One must be careful not to remove more than the peaks of the furrows, because the bearings must be tight on the shaft, with only enough give to allow preload adjustment. Enough material was removed (no more than 0.0001") to allow the bearing races to accomodate changes in the preload, but it's still a light press fit, only lighter. 6. Time for reassembly. Class 3 bearings are marked at the factory with the location of the high spots, and one is supposed to align the high-spot marks for lowest runout. This is described in "Timken Super Precision Bearings for Machine Tool Applications" (file "5918_09-09-29.pdf", 258 pages), in particular on pages 101 et seq. The marks are thick rings about 0.25" in diameter etched into the wider of the two flat faces of the bearing inner and outer races. The high-spot marks on the inner races had survived, but were scuffed off of both outer races, so it was not possible to fully high-spot align the bearing races. Why were the high-spot marks scuffed off the outer races? It appears that the root cause was that the quill nut was not tight enough, allowing things to move. This also explains the previously observed axial runout, where the spindle moved up and down noticeably in the quill. Axial runout was never measured, but appeared to be in the thousandths of an inch. 7. Tapping again proved insufficient to push the bearings onto the spindle, so pushers were machined from 6061 aluminum and cast iron pipe fittings, and the allthread was used to push the bearings back into position. This time, things were clean and the sliding surfaces were liberally slathered with "Mobil 1" synthetic grease (available from auto parts stores). The bearings were packed about one-third full with Mobil 1 grease before installation back into the quill. Millrite_Spindle_Saga_Photo5.jpg shows the overall setup for pushing the taper roller bearings back onto the spindle. Millrite_Spindle_Saga_Photo6.jpg shows more detail of the spindle nose area. The aluminum pusher is designed to allow the inner race to be pushed without involving the tapered rollers or the cage. Note that Grade 8 washers were required to prevent permanent dishing of the washers under the pressure required to move the bearings. The pipe caps upon which the washers rested in some setups were also faced off in the lathe, to ensure even loading. For periodic maintenance without disassembly, D.C.Morrison recommends injecting grease (not oil) through the pipe plug (item 12 on page 30), allowing the grease to work its way down the spindle and into the taper bearings as the spindle spins. 8. It proved difficult to get the spindle assembly back into the quill without the top ball bearing tilting and jamming in its recess in the quill. Used a shopmade steel washer and the black iron pipes and cast iron pipe fittings used for taper roller bearing pushing to push the top ball bearing straight into its seat, allowing reassembly of the quill. The instructions say on page 17 to be sure to install the snap ring "with the correct side up", but fail to say which side is correct. I installed the snap ring with the sharp edges up, on the theory that this orientation was most likely to allow the ring to stay in the shallow ring groove. The quill nut was tightened into the quill by hammering on the face pin spanner wrench with a lead hammer, to ensure that nothing will move. The quill nut will not bottom, instead being stopped when the outer races of the taper roller bearings and spaced are fully clamped. Recall that the quill nut has a left hand thread. 9. The spindle still has significant runout, but is better than before disassembly, probably because everything has been tightened. Further reducing runout will be the subject of another saga. Not shown is extensive cleaning with acetone and naptha to remove accumulated dirt and spooge from every nook and cranny. Joseph Gwinn, 19 October 2010.