A list I subscribe to has an ongoing controversy about the reason for
leveling a lathe. The majority opinion seems to be that this makes it
possible to easily test for skew in the lathe bed. In other words it's
not so important that the lathe be level, but that it should be
mounted on a flat surface.
The point then arose, what do you do on a ship? Lots of larger naval
vessels have machine shops. Even in drydock, how do they level their
lathes? Or do they?
I'm not a navy machinist. However, I was a shipfitter working on Navy
ships for eight years or so.
When a ship is put on a drydock for any extensive work, it is commonly
very close to level. However, the meaning of "level" isn't the same
as it is on dry land. What I mean is that if you rigged a water level
and compared the port-starboard marks layed out when the ship was built
to the water level, you'd be really close, probably within 1/16" over
sixty feet or so. But any given deck can be canted.
We leveled machines and machinery supports all the time. What that
means is level with respect to the ship as the frame of reference.
Of course, once the ship is on the open ocean, nothing is level.
In fact, the entire ship flexes and bends with the ocean. Any given
machine, however, is mounted to a machinery mount which is rigid
enough to approximate a perfectly rigid mount. Thus, a lathe on a
ship would be leveled with respect to its mounting platform. I
really doubt they do much high precision work out there anyway.
A whole lot of shipyard machine work is pretty coarse by uptown
(i.e. non-shipyard) standards.
John > A list I subscribe to has an ongoing controversy about the reason for
Lathes don't have to be level, what you need is to have no twist in the
bed. A level is a quick way to get in the ballpark, followed by test
cuts or the use of a precision testbar.
You can also use a length of ground rod in a chuck and find the average
reading of a dial indicator for 360 degrees of rotation of the chuck
close to the jaws, then repeat at some distance away from the chuck. If
the bed isn't twisted the average reading will be the same, although the
TIR will usually increase somewhat. The indicator point is placed on
center for this to work. You'll see this method referred to as "Rollies
Dad's...", and I tried to hit the site with this information, but it
seems the original sight is gone or there is a DNS failing somewhere in
the chain from here.
I understand that at least some shipboard lathes were mounted on three
legs to reduce twisting due to the vessel flexing.
So, something close to level is nice, as oil goes where you want it to,
but the real issue is removing any inherent twist in the bed and
securing the lathe to it's supports in a way that no undesired twist is
I was involved in installing some Ampex instrumentation recorders on the Norton
Sound. We were in drydock in San Francisco at the time. I went all over the
ship looking for a level. Secondary job was convincing people I wasn't sent by
a Chief on a jokester errand.
We just wanted these massive recorders to be approximately level with respect
to the ship. 1/2" wide tape, 60 inches/second with a frequency response up to
1 mhz. Pretty zippy stuff back then.
Mont Vernon, NH
I've got a scabby 4' carpenter's level with a long wedge of
pine screwed to one side. It came in an auction lot at a
boatbuilder (steel draggers, etc), and it took me while to
realize the wedge matched the angle of the railway in the
I spent eight years in the Canadian Navy as a marine engineer. I
liked machine work, so I became the guy who generally made small
parts, special fittings, etc.
The lathe I used was a nice machine, well mounted to a fairly stiff mount
and I'm sure it was levelled when it was installed. But leveling is
probably the least of the problems.
Alonside the machine was fine, but if work had to be done at sea it was
difficult. The workshop on my last ship, AOR - aulixiary oiler
replenshiment, was in a mezzaine in the engine room (common arrangement),
and if the CO started turning up the revs much past a gentle cruise the
vibration made it impossible. If he was late for a rendezvous and turned
it up to max during a cut you could count on having scrap.
John : A list I subscribe to has an ongoing controversy about the reason for
: leveling a lathe. The majority opinion seems to be that this makes it
: possible to easily test for skew in the lathe bed. In other words it's
: not so important that the lathe be level, but that it should be
: mounted on a flat surface.
: The point then arose, what do you do on a ship? Lots of larger naval
: vessels have machine shops. Even in drydock, how do they level their
: lathes? Or do they?
Department of Engineering Mathematics
My machine shop instructor was a navy machinist in the early 70's. He
told me that they would level a lathe by taking a series of cuts on
the outside diameter of a piece of round stock. Then measuring for
taper. Then adjusting the leveling bolts on the lathe. Then the
process would be repeated. I can't remember to whole jist of the
conversation, but he was saying something along those lines. I do
remember him saying it was a rather lengthy process. According to
him, it can be done.
You're not leveling the lathe so much as taking out or introducing twisting
stress in the bed to make the lathe cut with no taper. Having the lathe
actually level is a bonus, but not required. The lathe could be upside down
for all it cares, as long as the stresses are constant.
I can email you the relevant pages from the Machinery Repairman RTM
if you'd like. Glancing at it, it doesn't seem to say anything
substantially different from the reponses you've already received,
The purpose is only to to remove twist from the bed, using
1. The spirit level method, or
2. The test bar method (preferred).
If you need something "authoritative" to quote, email me (it might
take me 2-3 days to scan & convert to PDF).
This is pretty much correct. It's not at all important that the lathe be
precisely level, just that it be 'in tram' (properly aligned), that is,
turning 'true' without undue taper.
It happens, however, that getting the lathe level can be a good first
step in achieving a good 'tram'. Not necessary, but a good way to start,
Obviously, you can't get, or keep, a lathe or much of anything else,
'level' on a ship ... so other means of 'tramming' must be found, as you
This would be acceptable for all but the fussiest work, and in all but
the worst weather conditions. It can't be exact, however, for the whole
ship twists in a rough sea, so whatever the lathe is bolted to is not
precisely stable, and the lathe will also twist. Also, if the ship is
rolling or pitching appreciably, the lathe will be subject to
accelerations in directions OTHER than straight down (which is all it's
really designed to handle). At extreme angles you could also have
problems with lubrication, etc. Such perturbations, however, should be
slight except in the most extreme of conditions.
Probably correct in principal, but few lathes are designed to operate at
strange angles. Many lathe carriages are barely held on the ways except
by gravity. Lubrication would also be a problem, as the oil wouldn't
stay where it was needed.
Still, over some range of reasonable orientations, and and not too
violent motion, as on a ship, lathes can be adjusted to work perfectly
acceptably. A precise 'level' is NOT needed.
Stan Stocker email@example.com
That's interesting. The only three-legged lathe I've come across is an old
(circa 1880s) Putnam in the collection of the Charles River Museum of Industry
in Waltham, MA. (or was, 1996). The Putnams were made in Fitchburg, MA, where
I'm from, so I have some of them and follow/know the history of the company.
They did make Navy machine tools in the same era.
So maybe the one I saw was originally intended for shipboard use.
I have no other on point info.
A few years back I had a beater of a Colchester Chipmaster. It had an
all aluminum frame that had only three "feet" on it. I was told at
one time that it had come from a US Liberty ship. It was a very nice
beefy 10" machine, except that it had not been very well looked after
and there was terrible bed wear along the the tailstock sliding area.
It certainly can. It is the way I tram my lathes. It is the way Atlas
and South Bend recommended in their manuals too. It needn't
take very long to do it either if you use a dumbbell shaped piece
so you only have to cut at each end of the piece when you traverse.
This is called the "two collar" method.
Make a first skim cut, mike the collars, then if they aren't the
same diameter make a small adjustment of the lathe's "leveling"
screws. Then cut and measure again. The amount, and direction,
of the change in relative diameters tells you how much and which
way to make the next adjustment. 3 or 4 iterations should get you
down to tenths.