Levelling a lathe?

[Andrew TinsleyParticipant @andrewtinsley63637]

I am aware of how important this is to accuracy etc. My ML10 has now been stripped, cleaned greased and oiled and is looking a treat on its Myford stand!

The concrete floor of my workshop is far from level and the stand rocks alarmingly on two diagonally opposed stand legs! I DO mean alarmingly!

Now an old millright gave me a sheet of cork composite material many years ago. He said it was used under the feet or base of machine tools. Then when the raw bolts were tightened down, it was easy to level the working surface by tightening the raw bolts by the correct amount.

This seems to be a bit dubious on an ML10 and stand. I have no doubt you could get the bed level using a machinists spirit level. However the feet of the stand are only a little less than 2" x 2" and I fear the accuracy would soon be lost as the material settled.

I was thinking of making an area level, using self levelling compound, but this again seems to offer plenty of scope for error. What do members recommend to get my stand or rather lathe bed level? I am keen to get on with some urgent jobs, but refuse to use the machine in this state!

Thanks again,

Andrew.

[Andrew TinsleyParticipant @andrewtinsley63637]

[AjohnwParticipant @ajohnw51620]

In real terms you just need to stop it rocking. Getting the tray the lathe sits in levelled to the degree that can be achieved with an ordinary spirit level can be important if pumped coolants are to be used so that they will drain away. Also maybe to stop things rolling around.

Precision sensitive levels are used for checking machines not levelling them.

John

–

[JourneymanParticipant @journeyman]

You could get some adjustable feet like these from Axminster:- ***LINK***

If you dont want the lathe moveable, sink 4 lengths of studding into the floor and use adhesive to fix them. Use a nut and washer below each foot and a nut and washer above each foot and adjust as required. You could then fill in the gap below each foot with a concrete mix to make a nice looking pad.

Screwfix do a selection of resin fix concrete anchors ***LINK*** which would serve a similar purpose.

Cheers John

[Andrew TinsleyParticipant @andrewtinsley63637]

Hello John,

Now you have confused me! What do you mean by "precision sensitive levels are used to check machines, not for levelling them"? I can't see the point of checking them with a level and not doing anything about it (I.E. actual levelling them, if not level.) You might as well not check with a level if you are not going to do anything about it!

Sorry if that sounds a little aggressive, it is not meant to be, I am just confused by your statement!

I have read several lathe manuals and they always seem to start with getting the bed "level". The implication is that you cannot do accurate wok if the bed is not completely level. I assume that getting the bed level means that there is no distortion in the bed which would impede accuracy. I am sure you could get the bed distortion to zero with the lathe not being flat, but quite how you would do this is quite beyond me.

Andrew.

[Andrew TinsleyParticipant @andrewtinsley63637]

Thanks John (journeyman). Now that sounds a good idea!, Independent adjustment of each leg will get the lathe bed flat, without the fear of the packing material that I was I was considering, squeezing down with time. The cement would be good too but may get in the way of adjusting the nuts in a future resetting operation. I live in an area where we have had several earthquakes that have shook the house quite violently!!!!!!

Thanks,

Andrew.

[Roderick JenkinsParticipant @roderickjenkins93242]

Andrew,

The term "level" has developed a meaning of its own when referring to a lathe. When texts books refer to leveling a lathe they mean ensuring that the bed isn't twisted, which will cause the lathe to turn a taper rather than a cylinder. One way of ensuring cylindrical turning is by putting a precision level across the the bed shears at the the headstock end and comparing the reading with that similarly obtained near the tailstock. In an ideal world you would have the lathe mounted on a stand. The stand would be solidly fixed to the floor or be mounted on adjustable feat so that it does not rock. The stand does not need to be level with ground, indeed it would be beneficial to have some tilt on the stand so that the contents of the drip tray fall towards the drain hole. The lathe is then mounted on to the stand with adjustable feet, ensuring that the bed is not twisted. You can do this with a precision level. As I said, this is in an ideal world. Practically, you can mount the lathe on a bench or stand that is reasonably solid. Then undertake a simple turning test to see if the lathe turns parallel. Only if the lathe fails this test do you need to think about "levelling" the lathe. If you have a precision level by all means use it but trial and error will give you the same result: Adjust the lathe feet and repeat the parallel turning test until it is parallel to your satisfaction. The parallel turning test is the true indicator of what you want to achieve, the precision level just helps a machine installer to quickly do his job but, in my opinion, is unnecessary in the home workshop

HTH,

Rod

[JourneymanParticipant @journeyman]

Andrew, the concrete isn't necessary, it just makes it look pretty. I am not sure though whether you should install the cabinet/stand first and get that level and then add the lathe and do the parallel turning/levelling bit. Or if the whole job can be done in one go with the lathe already attached to it's stand. I think I probably lean towards a two phase operation!

Cheers John

[Neil WyattModerator @neilwyatt]

Posted by Andrew Tinsley on 27/08/2016 15:50:43:

Now you have confused me! What do you mean by "precision sensitive levels are used to check machines, not for levelling them"? I can't see the point of checking them with a level and not doing anything about it (I.E. actual levelling them, if not level.) You might as well not check with a level if you are not going to do anything about it!

A common misconception is that lathes need to be level – they don't (some are designed with angled beds). They do benefit from being fixed down without any twist in the bed and the engineer's level helps with this (it doesn't matter is both ends are angled at 0.5 degrees, for example, as long as its in the same direction), but the acid test is actually turning something or checking with a test bar.

Neil

[AjohnwParticipant @ajohnw51620]

The text book and various other sources confuse levelling lathes. Glad some one else explained why. I'm fed up with doing it so only post a short comment. What they really aught to say is make sure that the bed isn't stressed and twisted when the lathe is mounted where ever it's finally going.

Personally if Rod's method is followed I feel that there is need for a great deal of care when correcting turning errors by deliberately stressing the bed. Lathes can turn out for all sorts of other reasons and correcting those will twist the bed. It's better to forget the whole thing and make sure that a lathe bed is evenly supported and not distorted at all when bolted down. My usual way of doing that is to mount it on nuts and washers and adjust them with my fingers. Each one several times so that I can be reasonably sure that they are evenly supporting it, There also isn't any need to clamp the lathe firmly down with more nuts and washers on top of the feet so in my case those are just firmly finger tight. Do this and a lathe should turn true to very high limits. If it doesn't there are other problems. Worn or loose headstock bearings, worn bed or even a misaligned head. Over longer lengths a warped bed may even cause problems. I suspect in practice deliberately stressing a bed to correct a turning error went out of the window a long long times ago. It doesn't take much thought to realise the complications at the tail stock end when this is done but I suspect that the tailstock heights are still tolerance -0 + some tiny amount to help account for this practice on the assumption that there will be some error – a very very little one.

Actually I suspect many writings on bed levelling are cases of who ever wrote it not really understanding the process otherwise there would be a better explanation some where.

John

–

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Andrew Tinsley on 27/08/2016 15:50:43:

… I assume that getting the bed level means that there is no distortion in the bed which would impede accuracy. I am sure you could get the bed distortion to zero with the lathe not being flat, but quite how you would do this is quite beyond me.

.

Andrew,

Your assumption is correct … It's just that this whole process is littered with ambiguous words [or maybe jargon meanings assigned to ordinary words]

To compound matters, you have [I presume] used the word 'flat' in the sense of 'horizontal'.

.

I'm sure it is not 'beyond you' how to do this …

1. A decent spirit level has graduations either side of the 'level' marks
2. It is not difficult to fix a temorary spacer under one end of a spirit level

so, relying upon (1) and/or (2), you could do the whole job 'at an angle' if you wish

It is, however, usually more convenient to work to an horizontal reference.

MichaelG.

[duncan webster 1Participant @duncanwebster1]

For a start you could assume that if the lathe were suspended freely it would not have any twist (not guaranteed, but likely). Therefore put a long piece of bar in the chuck, and set a DTI against it at the tailstock end. If the DTI doesn't move when you tighten the lathe holding down bolts, you are not twisting the bed. You can probably tighten the ones at the headstock end to start with, then fit shims at the tailstock end to get no movement on the DTI. This should give you a good start for later tests.

[mechman48Participant @mechman48]

As long as there is no 'twist' in the bed the lathe can have some degree of out of horizontal…'level' what do you think would happen of fastening everything down solid on board ship with all the pitching & yawing going on in different types of weather & the effects on the hull.

George.

[AjohnwParticipant @ajohnw51620]

Perhaps how highly sensitive levels are used to "level" a bed might help. Say you put the level across the bed up by the headstock and note the reading and then did the same along the bed in several places. If the bed is flat the bubble will remain in the same position as it's moved along the bed. If the bed is twisted the bubble will move along it's scale.

Lets say that the bed is slanted 10 degrees front to back and 10 degrees along it length. There would be a need to adjust the level so that the bubble was near centre to account for the front to back slope as it's unlikely to function that well or even record that degree of slope. The slope the other way wont be a problem because the bubble will still be in the same orientation in the vial all of the time so will only move if the bed is twisted.

MichaelG will probably object to 10 degrees along the bed maybe rightly so depending on the design of the vial but the main point is that the check can still be made with levels that will show significant errors on an ordinary spirit levels. It will make no difference.

When level is being used in terms of machine tools it's more of a verb than anything else – it means to use the level. What they miss out is why and how. It's basically the easiest way to check a bed for twist and as they are analogue rather than digital devices they are extremely sensitive. We can easily eyeball very small bubble movements. Unusual but true compared with the sort of digital slope measuring device we are ever likely to own.

If you must fasten the cabinet down and need a simple way to to stop it rocking I would suggest a suitable size of items such as these. They explain the sizes at the bottom of the page.

**LINK**

Drill the holes for them in the floor. Discard the washers they come with and fit thick penny washers instead. Tighten up the nuts to fix them in place. Another nut and penny washer on the studs you now have to use to adjust and stop the cabinet from rocking. Then nut and washer to lock things in place. The feet may already have holes in them.

Or just pack it so that it doesn't rock. Lathes can fall over side ways as they tend to be very top heavy. That varies from one design to another. In my experience Myfords aren't too bad in that respect as they are pretty light lathes. A Raglan I had was something of a worry in that respect but would have still needed a pretty hefty tilt angle to cause it to topple over. I kept that on 3 broom stales so that I could move it around if needed.

Or if you don't want to fasten it down or order something drill holes suitable for 10mm bolts in the feet. Fit a nut to the bolt to use as a jack with a washer. Set it all up and another not and washer on the top of the feet to lock it all up once it's adjusted. It will be all resting on the heads of the bolts and wont move around very easily.

John

–

Edited By Ajohnw on 27/08/2016 20:30:15

Edited By Ajohnw on 27/08/2016 20:33:10

Edited By Ajohnw on 27/08/2016 20:40:18

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ajohnw on 27/08/2016 20:22:43:

Perhaps how highly sensitive levels are used to "level" a bed might help. < etc. >

MichaelG will probably object to 10 degrees along the bed maybe rightly so depending on the design of the vial but the main point is that the check can still be made with levels that will show significant errors on an ordinary spirit levels. It will make no difference.

.

Useful description, John

… and I have no 'objection' to your choice of an arbritary 10 degrees for the purpose of illustration.

'though obviously a precision level is most unlikely to register 10°

… which is why I mentioned (2) above.

MichaelG.

[Andrew TinsleyParticipant @andrewtinsley63637]

Thanks everyone for your input. I now have sufficient information to both fix the stand and level the bed of my lathe.

If I had thought for a second or two I would have twigged how to "level" the lathe when it was not horizontal!

No when I said plane I did mean just that. Not horizontal! I thought from my first post that I was making sure that my bed was horizontal (yes I did mean that this time!) Or shall we say that the bed was all in the same plane and NOT twisted.

When I first set up a lathe many years ago. I ensured the bed was planer and not twisted. I then went on to do the turning of a long bar and checking for constant diameter. In this case the bar was not of a constant diameter. I knew the bed was not twisted so I did not waste time looking for a non existent twist. The error was a very strange one, but that is another story.

Thanks everyone.

Andrew.

[Neil WyattModerator @neilwyatt]

One think people often forget is that to machine bed castings they have to be clamped to bed of a big grinder. It is quite possible for there to be residual stresses and there are other factors which need to be duplicated if teh lathe is to be the same shape as when it was made. If I recall correctly Myford 7-series beds were ground with a 1-thou 'hump' in the middle to allow for the weight of the saddle.

An idea for small lathes where the weight of the bed is unlikely to cause distortion is to only fix the headstock rigidly and mount the tailstock end on resilient pads.

Of course the ideal lathe bed is a single footed cantilever, but no everything is as well designed as an Adept

Neil

[Clive HartlandParticipant @clivehartland94829]

I have had my ML10 for 20 years now, it is sitting on a Myford stand and I use 4 rubber based threaded bolts.

It sits on a concrete and I basically level it end for end and then across, lock up the nuts on the feet and bingo. No special leveling and no problems.

The ML 10 is a short bed and is unlikely to get a bed twist unless some brute force is applied to any bolts screws etc. No need for grout bolts or bedding in at all.

If you had a lathe of 8" swing and a long bed then I could understand going the whole hog on leveling but the little lathe, set it down and use it, as whatever you do will make very little difference unless you force something too far.

[AjohnwParticipant @ajohnw51620]

I've got 3 sensitive vials Michael. One is on top of my digital one which reads to 0.1 degree. The vial is more sensitive so that the digital can be zero'd. Great fun as I had to keep swinging it through 180 degrees to check that it was on a level surface.

Another came off BW Electronics and is one of the military ones calibrated in some odd units. Much larger diameter and that one might be ok tilted rather a lot. I did buy that one as it was pretty cheap and might be needed one day and they can be rather expensive. I miss his "junk" listing as I have bought a few things off it.

Yet another – even larger diameter came with the Pultra. It wouldn't surprise me if that one was ok at 10 degrees at as described. I'm not sure what it was doing with the Pultra.

I don't believe the Myford myth Neil. More likely to be how they turned out. The best way to grind a bed lightly if that's being done is put parallels on the bed spaced as per the feet, stand lathe bed down on them and then grind the base of the feet. Invert with the feet where the parallels were. Clock up the bed and then regrind.

The other point on the myth is how often the saddle is in the middle ? Cutting forces with a rear parting off tool lift it. Normal cutting force push it down and out. The saddle slides on 7's are sensibly long. There are all sorts of holes in the myth.

John

–

Edited By Ajohnw on 27/08/2016 23:29:45

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Ajohnw on 27/08/2016 23:29:08:

I've got 3 sensitive vials Michael.

.

That's nice for you, John

MichaelG.

[Michael GilliganParticipant @michaelgilligan61133]

I'm still trying to visualise a 'Sensitive Level' that can also accomodate a 10° angle.

… obviously some clinometers and box-levels can do it, but that's cheating.

.

Just for fun … Here is a graphic illustration of what a 30° slope looks like

[I assume that most of us could mentally trisect that angle]

MichaelG.

[Mark CParticipant @markc]

Michael,

To measure +/- 10 degrees at .02/m you need a vial about 7m long (depending on how much measurement you want either side of 10 degrees! (Vial has an internal radius of 10m more or less for).

Mark

[Michael GilliganParticipant @michaelgilligan61133]

That's about where I got to, Mark

… But John's definition of 'sensitive' may differ from ours.

MichaelG.

.

Ref: …. "It wouldn't surprise me if that one was ok at 10 degrees at as described. I'm not sure what it was doing with the Pultra." [Ajohnw]

Edited By Michael Gilligan on 28/08/2016 11:42:41

[Mark CParticipant @markc]

Did you get as far as finding out the tube would be about 1200mm ID at the thickest section?

Mark

[Mark CParticipant @markc]

I suppose a 7m diameter spherical (bulls eye) with concentric markings would be useful for seeing exactly what was going on?

Mark

[Author]

Posts