Lathe alignment

[Rufus RoughcutParticipant @rufusroughcut]

My flippant reply was piece of paper being restrained from blowing away��,to which I was scornfully advised it was everything I could make from it, so I made a paper plane and missed the whole point of the exercise which was perception!

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Rufus Roughcut on 01/04/2013 12:51:02:

Hi Gents

As a novice I was asked what this is?????

With all guarded response and discriptions clear for a novice to understand what would you say ???

Rufus

.

Rufus,

Perhaps the question should have been "What is this not?"

… René Magritte should enlighten those who miss the point.

MichaelG.

[Mark CParticipant @markc]

After all the talk of mixed up names and misunderstanding and machining to millionths of an inch or any other submicron precision (and no takers for my leylines), here is a pictorial description that might help continue the controversy ….

 

I should have also drawn "bowed" as well but I have never seen a bed like that and it is too late to add it to the drawing now! (Bowed would be like sagging/hogging but rotated 90 deg.)

Edited By Mark C on 01/04/2013 18:48:02

[Ian PParticipant @ianp]

Mark

Useful reference diagrams. I am sure they will bring clarity to this confusing subject. You mention that you did not include 'bowed' because you had never seen a bowed bed, does that infer that you have seen all the others?

Ian

[Mark CParticipant @markc]

Ian, unfortunately I have owned a few, along with some other states of despair in the machine tool department!

Mark

[blowlampParticipant @blowlamp]

Mark.

Nice drawing there.

For this discussion we're interested in the bed on the centre right.

Placing an Engineer's Level at the left hand side of this bed should show the bubble to be close to the middle of the vial. Conversely, when moved to the right hand side, it will show that the bed is not level within itself by virtue of its twisted condition.

The Engineer's Level is just used as a means of removing the twist by comparing the inclination at both ends of the bed and making suitable adjustments to the jacking screws.

Another thing to remember is that you could also employ an Engineer's Level to help correct the deformities present in the two beds to the left of the twisted one.

Martin.

[Mark CParticipant @markc]

Martin, the interesting thing with all of this is that it would be hard to discern between any of these conditions by taking a test cut on two rings 6 inch apart! They would all produce a conical/tapered part. A lot of lathes will have 3 point levelling screws on each end which allows for hogging, sagging and twist.

You would always need some extra point of reference which is where the level comes into it's own – this is why they are used when machines are installed in a machine shop. The level is used as a comparator along the axis in two perpendicular planes to ensure it is truly flat/coplanar – it just so happens that the level used are very accurate and requires the plane to be close to level (plum) or you run out of divisions on the phial.

Mark

[blowlampParticipant @blowlamp]

The following is copied directly from Wikipedia and clearly uses the terms on which we're fixating.

An engineer's spirit level (or machinist level) is generally used to level machines, although they may be used to level large workpieces on machines such as planers. Spirit levels are also used in building construction, by carpenters and masons.

The upper image is a plain precision level used in the engineering field to level machines or workpieces, the lower image shows an adjustable precision level that has an accuracy of 1:10000. The adjustable nature of this level can also be used to measure the inclination of an object.

The accuracy of a spirit level can be checked by placing it on any flat surface, marking the bubble's position and rotating the level 180°. The position of the bubble should then be symmetrical to the first reading. Machinists levels provide screw mechanisms to center the bubble.

Some levels have a "vee" groove machined along the base which enables the level to sit on a round bar while remaining parallel with the bar's axis. They also have a smaller cross level to enable the second axis to be roughly checked or corrected and to insure the primary axis bubble is at the top of the vial.

A precision level (0.0005" per foot) is used to check the installation of precision machine tools in two axis. Lathes are manufactured with the base in a level plane and if they are not installed level distortions in the frame cause machining errors. Small milling machines are often roughly leveled but large mills are installed level. A worn lathe may have a twist introduced to the machines bed to ensure that it turns parallel spindle axis, by twisting the bed (that is worn) to the spindle axis. Lathes have 4, 6, 8 or more leveling screws. Usually, such a fix has limited utility for only part of the cutting tool (carriage) travel. Lathes usually require two or more leveling trials as the machine castings "settle in" to the adjustments.

Leveling a ship's lathe would be unnecessary due to the nature of the ships base – floating on water. However, the base of the lathe needs to be in the original manufactured plane to turn accurately and of course securely bolted to the deck. Very large machine tools are set on isolated concrete foundations and securely bolted to them to reduce vibrations and the possibility of movement from vibrations.

Martin.

[Mark CParticipant @markc]

Martin, If it says so on Wiki then it must be so…..

That page could do with being written by someone who really does know about the equipment and its use!

I am fortunate to have a pair of engineers levels (4 inch and 12 inch M&W) and a box level by H&W and it is surprising to see how susceptible a cast iron lathe bed is to even a small amount of load, let alone a bolted foot.

Mark

[Harold Hall 1Participant @haroldhall1]

Thanks Graham for your comments and I can understand your frustration when perfectly valid suggestions are not accepted. I can though understand the problem that a novice can have, not just with this thread but with threads in general also, when the answers given are so conflicting.

Harold

[blowlampParticipant @blowlamp]

Mark.

I don't understand your points.

I merely pointed out – albeit indirectly, that the Wikipedia entry uses the 'level(ed)' word in the sense commonly applied to machine tools, as that is what this topic now seems to be about – I never said it was the last word on how to use a level of any kind, although I see no glaring inaccuracies in its content.

Martin.

[Sub MandrelParticipant @submandrel]

I heard of a large machine tool that kept going 'out' until they discovered it was only spot on when the tide in the nearby river was at the same state as when the machine was set up.

I also heard of a large bed machine so long it had to be levelled by optical means as otherwise it curved with the curvature of the earth.

Neil

[Mark CParticipant @markc]

Martin, given that I read that entry twice before deciding it was tosh, I just read it again and have to take a completely different understanding to the first two times! I originally thought the author indicated that machines will warp if they are not reorientated as they were made/cast (in space). It highlites the problems with "written only" descriptions though.

With regard to machines on ships – I have no idea how they are adjusted but I would be inclined (ha ha) to use an optical (laser?) method.

Mark

[Mark CParticipant @markc]

Neil,

I have had the pleasure of visiting the Airbus wing factory a time or two and they have some kit there that runs on rails set in the floor that is quite accurate (probably better than most of our lathes) and that is set using laser alignment tools due to it's size – it is big enough to take a whole wing! I just worry now that I might be on a plane that had wings being made when we had all the snow on the ground….

Mark

[Harold Hall 1Participant @haroldhall1]

Mark C

As Graham stated earlier "it is possible to to carry out the levelling of a lathe with a Dial indicator". I covered the method in my book "Lathework a complete course"

Harold

[Mark CParticipant @markc]

Harold, over the whole working length?

Mark

[Michael GilliganParticipant @michaelgilligan61133]

This is worth a look

MichaelG.

[jason udallParticipant @jasonudall57142]

Some times , to understand a problem it is better to understand the opposite/ extreem , then see how it relates.

If a lathe is put into use with its rotating axis mostly horisontal ( this being a horisontal lathe) but chocked up front to back 1 " at front headstock end and 1" high at rear tailstock end.

IF the lathe is perfectly ridgid ( inelastic)…then it would behave as intended.

If the lathe is elastic … lathe will show some "warp" from one end of travel to the other andtool path will follow a curve headstock to tailstock. and even cross slide.

so far so obvious.

If the weight of the lathe is DESIGNED to be equally spread over the two ( say) feet at headstock end and a lower portion at the tailstock again equally… We can fix the lathe to a "light" plane and examine from afar…there will theoricaly be a point where the plane can be balanced on a single point…

NOW your lathe is LEVELand at equilibrium, even if the ways are not…. and it is in this attitude that the lathe would be most stable and least stressed..

Now if our plane is horizontal NS EW and parallel to the lathe ways / spindle axis a ready way of reaching this happy state is provided…

It now requires the manufacturer to build his product like this…..

So who is to tell? Maybe your lathe would perform better with a little "run" on the ways

CNC manufactures / installers have a vested intrest in providing machines that do the job.

I have seen many lathes fitted by the makers staff and only HARDINGE don't fiddle about..bolt it to the floor so it doesn't walk yes..level so the suds / stock doesn't fall/run too much through headstock..yes.. But then again thats hardinge..( machines are one big "casting" of resin concreat)

MAZAK do the twist/cut/twist/cut dance ( often chasing 1/10 MICRONS)..( never seen more than 300 mm spirit level used for "leveling" )..

My opion?.. level it via the ways, if no other datum available, as first guess.
It also stops that drill you "just put down" rolling off into the darkness……

[Mark CParticipant @markc]

Gray, I think you would run into problems with the bar and DTI method. First, you need to check in two planes – vertical and horizontal otherwise you do not know if it is twist or hog/sag. The second problem is the self weight of your bar, which will sag due to it's own mass. Third, you will never get a fully fixed connection in the chuck jaws – and it does matter in this instance as it will add to the self weight sag measurement of the bar. The other problem is getting on top of the bar, with a 30mm bar you need to be bang on in both positions relative to the bar centre line otherwise you will get errors due to the curvature of the bar – this is unavoidable if the bed is not true and here in lies the contradiction that for it to work correctly you first need a bed that is flat and a bar that is not affected by gravity.

On the subject of laser alignment, I do not see that as particularly onerous (without knowing the precision required – nanometer positioning would be starting to get interesting!) given that the light source is fixed and not moving (do I understand correctly?). I spent time working with the optics guys at Physik Instrumente in Germany and they do this sort of stuff daily. I would have used the term "static hydraulic ground pressure" related to the wet sandy ground under your theatre, rather than liquefaction – which is quicksand to everyone else! The structure should not have been affected if it is a box form and was carefully designed for the environment, again I have been involved with the installation of lead shielding on some fairly accurate kit and this is always taken account of in the building design which is generally undertaken by structural engineers with experience in this area.

So, to conclude, it seems to me that if you want to spend hours of your time messing about with a second best solution that might or might not produce the flat plane you desire, then feel free. If you want to get the job done, minimum of fuss and with the greatest ease, then get a level or have it done by someone who makes a living doing it. In the end, there are so many other things that people can spend their time fixing and perfecting that levelling a bed to these accuracy are probably a waste of time especially as most people will never need to turn long items at great precision anyway (depending on the job, there are probably better suited machine processes to do it as well). Perhaps the answer is to borrow a level or get a friend who has one to help out. Of course, if you have a level and the machine is set plum then you can also use the level to set work up quickly as well!

Mark

[Michael GilliganParticipant @michaelgilligan61133]

Mark,

I have sent you a personal message, regarding H&W

MichaelG.

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

(1) Most modern CNC machines are of superb accuracy anyway but some of the best have active error correction ( various names for this ) as well . This means that the controlling software can make continuous minute corrections for any inaccuracies found by initial testing or by experience .

Interestingly this is the same basic process as used in top of the range traditional type jig borers with their hand filed correcting cams and is the same process as used by an old time machinist when he applied minute biases to machine slides at different places as the cut progressed .

On non CNC home workshop machines fitted with DRO but having minor misalignments it would be possible to ' map ' the errors by testing and set out a table of local corrections to be applied by the user .

On a CNC home workshop machine it would be relatively easy to extend this idea to complete mapping of errors in machine and automatic correction – going some way towards system mentioned in opening paragraph .

We then have the interesting paradox that an inaccurate machine with error correction might be more accurate than a conventionally accurate machine without error correction !! .

Some machine control systems for home workshop use may have facilities for active correction but apart from backlash correction they seem to be little used .

(2) There are many alternative ways available now for aligning machine tools . Most of these are optical and some of them would not be too difficult to set up for home use . We can start a separate discussion on this if anyone is interested .

Michael Williams .

[NJHParticipant @njh]

Michael

"We can start a separate discussion on this if anyone is interested ."

I think that would be a good idea. This topic is now a thousand miles away from Jan's original request for advice on adjusting his Myford lathe! I trust he has taken the advice he was given initially but , if not, he is probably now contemplating giving up model engineering and transfering his interest to something less controversial. (Like politics perhaps!)

N

[jason udallParticipant @jasonudall57142]

As always ( it seems) some of these disccusions become " religous" and the OP will often have lost intrest and gone off to try for them selves…

Michael Williams .. new thread…great idea… can we duplicate this one and rename? (moderators ?)

As to self "correcting" cnc well I think thats what is called closed loop control.

I have designed and made subsystems to measure the PART as its made and allow for taking further correcting cuts…also correcting for thermal growth and tool wear into the bargin….as I say at the end of the day its the shape of the part that counts…

( as an aside…. ever used a laser mike?..what fun those are…try explaining to sales rep why its measureing the major /minor axes of an ellipse not the diamter of the rod…he he)..

anyway back to the model world…

A leveled ( think of balance scale in equilibrium ) might not be horizontal but should be in balance and under minimium stress…this is a good thing..how you then need to warp it to make it cut straight is another matter..

What ever allows you to turn parallel when desired and face off square ( yes I have heard the rumours and myths about normally setting this to turn concave ..pardon me but I will aim as default for square.. and profile as the job requires)..the two secondary functions of a lathe..is fine with me….anyone guess my primary function for lathe?.

.

[Mark CParticipant @markc]

Gray, Simply commenting on what has been raised and written here, did not know there was a history to be read first – perhaps you could point me in the direction of the setup description and I will oblige by reading it. I am not bothered what you do with your time or equipment, but setup does seem always to prove emotive along with a number of other subjects.

The most important thing is to have a method that suits your requirements (in all things), it is simply helpful if you understand the options and the implications regarding their use…. to this end, perhaps a full description with pictures (they do save a lot of problems) of the different setup methods along with their merits and pitfalls and suitability to different types of work would be more helpful than continually spiralling into alignment issues at the sub atomic level (who cares when you are fixing a garden tap?).

Mark

[Mark CParticipant @markc]

Gray, can you turn that into a link please, there does not appear to be anything under the 79997 number and the link is just text?

Mark

[Author]

Posts