Leveling a lathe, why ?

will

Flashlight Enthusiast
Joined
Apr 14, 2004
Messages
2,597
I realize that you want the machine to be reasonably level due to fluid drainage. Also you want it supported on all the 'feet' to eliminate any possible twist. I can see it if you have a setup like a combination lathe/mill and you use a level to compute angles for milling.

but - how accurate does the level on a lathe have to be ?
 
how accurate does the level on a lathe have to be ?
It isn't really critical ... as long as both ends are within one vial graduation (.0005"/foot) :D

Here's one of my bookmarked pages, that says it well:

Because:
  • A properly leveled lathe is the first essential for accurate work and long service life.
  • The built-in accuracy of the lathe can be permanently destroyed by improper leveling.
  • Satisfactory performance is impossible if the lathe bed is out of level as little as one thousandth of an inch.
AN IMPROPERLY LEVELED LATHE WILL

  • CHATTER
  • TURN TAPER
  • BORE TAPER
  • FACE CONVEX OR CONCAVE
  • SCORE BED AND CARRIAGE WAYS
  • SCORE SPINDLE
  • RUIN SPINDLE BEARINGS
  • MAKE CARRIAGE BIND
  • TWIST HEADSTOCK AND SPINDLE, BED, CARRIAGE AND TAILSTOCK OUT OF ALIGNMENT RESULTING IN EXCESSIVE UNEVEN WEAR.
http://www.atlas-press.com/tb_bedlevel.htm
 
but - how accurate does the level on a lathe have to be ?


First off the lathe does NOT have to be level.
However most of the time we do set it level so that oil does not pool in the wrong place in the chip tray etc,etc.

When people say that a lathe must be level, what they really mean is
" I need to set up my lathe as close as possible to the way it was in the factory where the bed was ground.So there must be no twist in the bed".

Where to position the level is the next question. There are a few answers on this.

I like to set my level on the cross slide ,90° to the ways and then wind the saddle back and forth. If you have no movement of the bubble, you are in the exact setup as when the bed was ground. (You might have to shim one end of the level so the bubble is close to dead center before attempting to take a reading.)

Now you should be able to turn two 'collars' on a 4" test piece and both diameters should be equal / close.

Now don't expect perfection here unless you are fortunate to have a $50,000 Hardinge.
 
Last edited:
Thanks.

Reading through the leveling instructions it appears that the machine is being bolted to a solid base. If it is not level, there is the possibility of putting in a twist. With many of these small benchtop lathes, I would think it would be very difficult to twist the bed.
 
With many of these small bench top lathes, I would think it would be very difficult to twist the bed.

Basically you have a rectangle with 2 bolts on either end. Now the base you are going to bolt to might be absolutely flat to within microns but its flatness might not duplicate the position of the lathe when it's ways were being ground.

Set up a level as I describe and then tighten down the 4 bolts and I guarantee
things will move and you will finish up having to shim one of the four bolts.

A small lathe takes just as much time to set up correctly as a big guy.
The principle remains the same.
 
When I was working as an apprentice industrial electrician I wired up a few lathes that were really big. I didn't pay attention to their exact size but they were easily 25-30 feet long. While we were running conduit and wire the riggers were busy leveling tweaking and anchoring them down. Again I didn't pay too much attention to them but I remember hearing them talk about how it was extremely critical that they be as close to perfect as possible. As we got them wired up they would come back and check them out again for level and settling and tweak them some more and run them back and forth and make sure every aspect of them was perfectly level. when they were done with all the tweaking then they back filled around the leveling hardware with something like stucco/mortar or concrete. I wouldn't have imagined that something that huge wound be flexible or even need to be level.
 
Last edited:
It's amazing where precision takes over from self supporting.

most materials and machinery will support there own form quite well, but to function properly takes that precision adjustment to eliminate misalignment and binding.

It's a good idea to have a level shop just because oils and round pieces won't tend to go where you don't necessarily want them.

Like said before the proper term would technically be to keep your lathe flat. And as flat as possible.
 
My Enco 7x10 lathe has an attached drip pan. This is held on by 4 rubber feet. This little lathe gets moved around when I need bench top space. The table top is level, the rubber feet keep the machine from sliding around.

thanks to all the answers here..
 
With many of these small benchtop lathes, I would think it would be very difficult to twist the bed.
The smaller the machine, the easier it is to twist the bed. There is little mass relative to the small forces involved.

Twist is a relative term. With a lathe, having one end in a flat plane & the other end at an angle to that flat plane is a twist ... even though the level bubble indicates "only" .0020"/foot. That small amount of twist will turn a different diameter at the chuck & at the far end of a part, and will bore a different diameter as the boring bar advances toward the chuck.

A master precision level will read to 8 arc seconds/foot, which is equal to .0005"/foot. If the bubble reads four graduations different from one end to the other (.0020"/foot) you'll never get accurate results, especially on parts over an inch in length. One graduation is acceptable, less than one is better.
 
I realize that you want the machine to be reasonably level due to fluid drainage. Also you want it supported on all the 'feet' to eliminate any possible twist. I can see it if you have a setup like a combination lathe/mill and you use a level to compute angles for milling.

but - how accurate does the level on a lathe have to be ?

OK, a lathe doesn't have to be level - it has to be twist free. The easiest way to do this is to make the lathe level. If you bolted your lathe to a deck of a ship with no twist, it would not care about the rock and roll of the ship at all (assuming the ship isn't bending, and they do)
 
...snip... when they were done with all the tweaking then they back filled around the leveling hardware with something like stucco/mortar or concrete. I wouldn't have imagined that something that huge wound be flexible or even need to be level.

It's called grouting, and is done all the time in industry

Yes, even big lathes bend, heck, an HLVH or a 10EE (or other super heavy duty short bed toolroom lathe) will bend. How level do they have to be? Depends on the class of work you want to/need to do

Not a lathe, but I met a guy once who showed me pictures of a shop that used to scribe master defraction gratings. Now these things have to be accurate to a fraction of a wavelength of light. TOUCHING a machine meant that it had to temperature stabilize for an addition 2-3 days. The shop was underground, one machine - the machine drew something like 100 watts when running - in the room was a 100 watt lightbulb. After the doors were shut, and the room stabilized with the lightbulb on, when the machine turned on, the bulb went off, machine off - bulb on

The guy I met had an interesting job there - he wrote the software that could predict which ball in the ball bearing was in what spot relative to both races, and compensate for the errors - and they were buying the best ball bearings made
 
It's called grouting, and is done all the time in industry

Yes, even big lathes bend, heck, an HLVH or a 10EE (or other super heavy duty short bed toolroom lathe) will bend. How level do they have to be? Depends on the class of work you want to/need to do

Not a lathe, but I met a guy once who showed me pictures of a shop that used to scribe master defraction gratings. Now these things have to be accurate to a fraction of a wavelength of light. TOUCHING a machine meant that it had to temperature stabilize for an addition 2-3 days. The shop was underground, one machine - the machine drew something like 100 watts when running - in the room was a 100 watt lightbulb. After the doors were shut, and the room stabilized with the lightbulb on, when the machine turned on, the bulb went off, machine off - bulb on

The guy I met had an interesting job there - he wrote the software that could predict which ball in the ball bearing was in what spot relative to both races, and compensate for the errors - and they were buying the best ball bearings made

That is wild.
 
Mass doesn't prevent twist; cross-sectional area does.
The machines built for the highest precision, like toolroom lathes, use a combination of large mass plus extensive ribbing & cast in cross bracing (additional cross sectional area). You're right that mass doesn't prevent twist, but it does serve to dampen vibration.

A good read on controlling machine vibration (which has nothing to do with leveling a lathe:)) is in the current issue of Fabricating & Metalworking:

http://www.fandmmag.com/print/Fabricating-and-Metalworking/Controlling-Machine-Vibration/1$3022

On the second page, look at the box titled "A Brief History Lesson". One of the best summaries I've seen on box ways vs. linear guides - all current CNC machines use one or the other.
 
Top