What's new
What's new

Replace babbitt split bearings in old wood lathe headstock

Joe Michaels

Apr 3, 2004
Shandaken, NY, USA
I am glad to read your post. You are correct about preheating the bearing shells prior to babbitting. We preheat with either a 'rosebud' (large heating tip) on an oxyacetylene torch, or with a propane weed burner. For something like the wood-lathe spindle, if we were to roll the clock back over 100 years, babbitting would be done as follows:

-I notice the shells have keyways to mechanically lock in the babbitt. Were the shells simply half-round bores, reliance for holding the babbitt in place would be placed
on the 'bond' between the babbitt and the cast iron (same as 'wetting' in soldering). Some babbitting was done by pre-tinning the iron shells. This required using a
flux and a 'tinning compound'. SOmetimes, pure tin was used for this. Areas of the bearing not to be babbitted might be painted with 'whiting'- a water/chalk solution.
Anything with water had to be 100% dried before pouring babbitt into or onto it. Preheating for tinning and leaving the shells hot would insure this was the case.
-set bearing shells up in the headstock, or in a simply jig on the bench so the bearing shells are in alignment, and provide some means of centering the spindle (or
a dummy shaft known as a 'mandrel') in the bearing shells.
-coat the journals of the mandrel (or the actual spindle) with 'lampblack' (carbon soot). We do this most often by lighting an acetylene torch and not turning on any
oxygen. The rich flame puts plenty of soot on the journals. Oldtimers used a kerosene lamp to do this, a bit slower but it works.
-assemble the spindle (or mandrel) and bearing shells so the journals are centered in the lower bearing shells. The ends of the shells have to be 'dammed' to
contain the molten babbitt. Sheet metal 'washers' and clamps are one way. No matter how tight you think you have these sheet metal washers, molten babbitt often leaks out. This is stopped by using a putty made for the purpose, it's name being something like "Babbitrite" or "Dambabbitt". The putty is re-usable. It contains no
water and has a high enough flash point that it does not catch fire when the hot babbitt hits it.
-preheat the assembled bearing shells/spindle (or mandrel). We use an oxyacetylene torch or 'weed burner'. The oldtimers had to contend with preheating the shells
in a forge fire, or gas fired furnace as used for heating soldering coppers. Very scientific means of checking heat: spit on the shells and if your spit beads up and
really sizzles, things are nearly there. Heat a bit longer.
-melt the babbitt in a cast iron pot (the kind plumbers used to use for lead work). Again, real scientific means of checking babbitt temperature: take a piece of clean
pine wood and touch it to the surface of the molten babbitt. If the pine just smokes, the babbitt is hot enough. If the pine bursts in flame, you are screwed. Leaded
softer babbitts do not like being overheated. You touch the pine stick to the molten babbitt as soon as it is fluid in the pot.
-stir the babbitt with a steel rod you have cleaned with emery cloth and made sure to degrease. Preheat the rod.
-pour the babbitt using a ladle with a pour spout. A plumber's lead ladle is all that is used. Pour from one side of the bearing shell. You can preheat a thin steel rod
(we use stainless steel TIG welding rod), and work it in the molten babbitt in the shell. This will work up any trapped air. Be quick about it as the babbitt will want to
freeze quickly. On the thin shells you describe, the rodding is likely neither possible nor needed.

When the babbitt in the bottom shells has cooled, take a coarse file or rasp (I use a farrier's hoof rasp for this), and take the babbitt down flush at the split joint of the bearing. Finish the split joint surfaces by using a straight-edge scraper. A piece of broken power hacksaw blade ground to a nice straight edge works great for this.

Coat the freshly babbitted lower halves of the bearings with lampblack (carbon soot), re-coat the spindle (or mandrel) and reassemble in the jig on your bench.
The upper shells may have holes drilled in them for babbitting, aside from the oil holes. Clean these upper shells as noted, above. If there are extra drillings for babbitting, great. If there are, you can plug the oil hole with some of that babbitting putty. If no separate babbitt holes, you will be pouring thru the oil holes.
If that is the case, you will need to make a pouring basin or some means of funnelling the molten babbitt into the oil holes.

Make some shims out of sheet metal and smoke them thoroughly. Assemble the upper and lower bearing shells with these shims in place. Shims must contact the shaft journals. Making the shims with slotted bolt holes and oversized so you can tap them to drive them hard against the journals is what I've done on similar work.

Dam the ends of the bearings, same as the lower halves.

Preheat the whole assembled bearing and pour the babbitt into the upper bearing shells.

When cool, open the bearings. If you got it right, the bearings will show a nice clean surface with no porosities (from air bubbles) and no cold laps from areas of the babbitt chilling and freezing before more molten babbitt arrived to meet it. On larger bearings, having another person with a second ladle is how we do it. Preheating and keeping the heat on the bearing shells up to and during pouring (on larger work) is also how we make sure we do not get chills and cold laps in the babbitt.

Getting back to the lathe headstock as it is. I would first attempt to remove shims and and 'blue in' the spindle journals. Using a thin coating of Prussian Blue (a paste type dye) on the journals and rolling them in the bearings will show the contact pattern. If contact is minimal, I'd try scraping the bearings to improve contact.
You do not need 100% contact, particularly on something like this old wood turning lathe. A pattern of blue spots well distributed over the bearing surfaces, say about 80% or better is going to be more than sufficient for this application.

I use a 'spoon' type bearing scraper, well stoned, to scrape in this type babbitted bearings. After each iteration of scraping, I take a straight scraper (made from a piece of broken power hacksaw blade) and give the bearing surface a very light 'shave'. This takes down any burrs or whiskers of babbitt. I then wipe the surfaces with a clean rag or paper towel and solvent (mineral spirits). Re apply the Prussian Blue to the journals an roll it in again... inspect... repeat the scraping touching only the areas where the blue appears. Sometimes, you will see what I call 'leopard spots'. These are areas of blue with a small bright metal center. These are spots where a real hard contact with the journal was made. The bluing was squeezed away from the center of the spot. When you see these leopard spots, you scrape the center and leave the bluing alone. COunterintuitive, I know.

If you get a good contact pattern with scraping, scrape a nice little chamfer on all edges of the bearings and assemble with shims. As for your client wanting to run the spindle 3000 rpm, he better have his hospitalization and life insurance paid up. That wooden headstock pulley will fly apart at that kind of speed. For that matter, wood turnings, unless small diameter and very sound wood, will also be apt to fly apart at that kind of rpm.

Assuming he replaces the pulley with a steel or aluminum pulley, then he better plan on having the spindle/pulley dynamically balanced for that kind of rpm. Otherwise, the lathe will want to dance away. For that kind of rpm, the type bearings you show are not up to the job. Those were intended for slower speeds, maybe several hundred rpm.


Nov 19, 2017
Greenville, SC
Thanks once again for your input.
The guy said he is an engineer and used to work with machine tools before he retired, and was used to very high accuracy.
In all honesty I was shocked when he told me he clocked 0.001" TIR, and was concerned.
When he started talking about 3,000 rpm, he lost some credibility with me, and I never even thought about the pulley flying apart, duh.
I was obviously trying to be nice and reasonable in my attitude, in case there might be some work in the future, but on reflection I'm thinking he'd be a total pain to satisfy.
Thanks again,


Feb 22, 2006
Smithfield, Rhode Island
The likelihood that lathe was intended to run at anything like 3000 RPM is nil. I doubt a Babbitt bearing would support close to that speed, at least not with pressurized oil...certainly not oil drip cups.


Dec 21, 2012
Brisbane Qld Australia
once I spent a lot of time with an antique wood lathe for a friends father .......about a month later ,he turns up with the end of the spindle hacksawed off,and wants it fitted to the Chinese woodlathe hes just bought......I was not a happy chappie.


Jun 1, 2004
i would keep the Babbitt type of bearings but i would start off by regrinding the spindle so everything id true then re pore the bearings if cared for will last a life time or two and if you can not do it find someone who can like somebody that dose mod. t's or A's and if there are no oilers then add some


Jun 16, 2001
St Louis
The likelihood that lathe was intended to run at anything like 3000 RPM is nil. I doubt a Babbitt bearing would support close to that speed, at least not with pressurized oil...certainly not oil drip cups.
Depends on spindle diameter... surface speed. Small spindles need higher speed in rpm to build up an oil wedge.

What about 2 pole motors? Lots of them running with plain bearings, at 3450 rpm under load, nearly 3600 at no load.....

it's not necessarily unrealistic for the spindle. The pulley may be different.