|Anthony Salisbury||29/09/2009 23:10:45|
|32 forum posts|
I have to admit this one is slowly driving me mad! In long / short when turning on my lathe I have a consistent runout / taper on all my turning.
For example 1" Brass turn to 20mm over 50mm. I turn perfectly 20mm approximately 50mm away from the chuck but when checked near the chuck it is approximately 20.1mm.
I have clocked up the spindle with a parallel bar and I have approx 0.02mm out of round which isn't bad. Runout over 130mm is approx 0.015mm again can't grumble.
I'm mounting in a three jaw chuck.
Anyone please help all ideas welcome! It is literally driving me mad. I'm used to having a lathe that will generally turn parallel!
Thanks for reading
Edited By Anthony Salisbury on 29/09/2009 23:11:29
|451 forum posts|
A while ago I had a problem like Yours with my Emco Compact 8.
Please read "Lathe work - A complete course" by Harold Hall.
In page 37 You will find a recommendation which saved me the day.
Amongst a wealth of technical and practical information information He says
"Typically, if the teste piece is larger at the outer end, the front mounting at
the tailstock end requires to be raised, or if smaller, then the mounting requires
to be lowered, etc. etc.".
I hope this helps You.
|1008 forum posts|
You have done the standard turning turning test on 2 collars and shimmed under the feet of the lathe accordingly?
You have runout on your spindle? Seriously? Detectable runout with a DTI against the inside of the morse taper, or on the Camlock taper? (Not on a bar held in a chuck of course which will runout all over. <2 thou runout is good for a good 3 jaw, and I bet our cheapo Chinese 3 jaws do a lot worse than that.) Bearings checked and adjusted?.
You should have sensibly, zero runout.
Mounting in a three jaw chuck shouldn't affect trueness. That doesn't mean a 3 jaw will hold truly concentric - it wont, but if you turn a dia (ie skim a surface) that should be true and not runout, as should any other circumferential point on the same bar machined at the same setting. What you can't do is clamp a test bar in a 3 jaw and draw any conclusion as to the truth of the lathe. You can set it to run dead true in a 4 jaw if you wish, and then clock off that, but not a 3 jaw.
Checked the tailstock is aligned if you are using that to support your bar. Very occasionally the tailstock may be low or high, but more likely it will be out side to side.
The way to do the adjustment is to do the standard tuning test and shim till that comes out right. Then put a test bar between centres head to tailstock, and adjust that out. Wise to use a soft centre in the nose of the lathe and skim that new and in situ before doing the test.
Tools are good and sharp, and you are not measuring after some mega deep cut? Doesn't sound like it if the large dimension is nearer the chuck.
Are the chuck jaws allowing flex, even though hte lathe is straight? That would normally give the larger dimension further out, unlike yours.
So do the tests and see what happens. My 12/37 was a mile out plonked on the workshop floor, as delivered. Properly adjusted, and I just reset it the other day, its turning parallel to better than .025mm in 12" They do move about a bit, and I do mine every sort of 6 monthly or so.
Edited By meyrick griffith-jones on 30/09/2009 00:01:32
Edited By meyrick griffith-jones on 30/09/2009 00:06:42
|1008 forum posts|
Dias - that's the standard shimming test. Myfords have rather nice screw jacks for the purpose, but most lathes needs shimming.
Turn 2 collars - sharp tools small cuts, preferably in aluminuim, on a rigid bar (1 1/12" approx depending o nthe lathe) say 6-8 " apart or more depending on the size of the lathe. Distance is not not critical. same setting - and don't remove form 3 jaw.
If the diameter further from the chuck is the smaller of the two diameters, then you need to move the work AWAY from the tool. So shim under the front mount, tailstock end. If its the larger dimension, you need to move it towards the tool - so you shim under the back mount, tailstock end. Headstock end is not moved. Eventually you get the 2 collars the same size and your lathe is set right.
Edited By meyrick griffith-jones on 30/09/2009 00:15:59
|Anthony Salisbury||30/09/2009 09:55:50|
|32 forum posts|
|Anthony Salisbury||30/09/2009 09:58:48|
|32 forum posts|
Do you get the 0.025mm over the 12" supported by the tailstock? What about just mounted in a three jaw chuck? Can you get as good tolerence say with a 1"bar over 300mm not support by tailstock?
Thanks for the great replys
|1008 forum posts|
Ordinary turning there is no way I'd not use a rotating centre for something 12" long. Then because of the runout in a live centre, I'd change to a dead centre for the last 2 cuts. And if it were thin, then i'd be using a travelling steady as well.Its far too much overhang and it is bound to spring. So its not a valid question, if you'll excuse me putting it that way
But yes I do get that sort of lack of taper because both tailstock and lathe bed are properly adjusted. In fact on the Myford, currently the taper is 0. (but that won't last!)
The point is that you cannot take a bar and simply put it in a 3 jaw and expect it to run true to critical limits.
Towever if you TURN a diameter on it then that diameter will run true (and you cut the excess off having found your sculpture within the bar as it were)
So stick a rigid bar in the 3 jaw. Rough turn 3 hollows, so you have 2 widely spaced collars. (With a rotating centre installed if it were me) You want a cotton reel shape.
Remove the centre and now lightly skim the collars to a good finish using power feed, WITHOUT moving the infeed, so you are testing the alignment of the bed.(Hence the ali and light cuts) ABSOLUTELY at the same setting.. They will indicate 0 runout. Now skim them again (.02mm), without altering the infeed inbetween and measure with a MICROMETER. If the one further from the chuck is bigger you need to move the toolpost towards the work - not by using the infeed, but by putting a shim under the front foot of the bed. If the outer collar is smaller, its a bit of shim under the back tailstock foot to move the tool away. Reskim remeasure, readjust. until its right. A good game played slowly, buty once it is right, any future adjustments you can make by a nipping up on the holding nuts usually. I've never had to add extra shims.
The assumption is that the headstock and hence axis of the bar/collars remains still, and you twist/straighten the bed by moving the tailstock end. (Not the tailstock but shims under the mounting points at the tailstock end, so the whole bed is moved. With a lathe like a 1224, you will need to slacken the mounting bolts at that end , and then with a hydraulic jack, lift the bed a smidgen, slide a shim under, drop the bed and retighten and then recheck). That in turn moves the saddle relative to the CL of the bar or spindle bore and takes out any taper. If you have a 24" lathe and you are working at the 12" point, then very crudely, if the difference in diameter is .001", the difference in radius is.0005. You are at the halfway point, so you need about .001 of shim. And if you are working 1/3rd of the way down the bed, then pro rata. Thats just a very rough starting guide and allow a whole evening for the exercise!!!!!!.
You can do a rough check with a test bar mounted in the morse taper (NOT a 3 jaw) and a DTI in the toolholder but you do have to finalise by cutting collars.
Once you have the lathe straight, then you can put a test bar between centres and clock on it, to take out any tailstock misalignmant. Once you have done that, you will find that you can set something to run true in the 4 jaw, put a clock in the end of the bar and run the tailstock centre into place, and you will get no movement on the clock
When you check in the future, you can certainly put the same bar in the 3 jaw, but you must skim the collars true and then take the cuts.
The actual diameter of the collars is unimportant - its the DIFFERENCE that you are taking out.
I may be teaching my Granny, (and I aplogise if so) but I think there is a little confusion, and perhaps also a touching faith in what a 3 jaw self centring chuck will do .
Runout is radial misplacement. - hold something in the 3 jaw that is truly round, and the clock will indicate movement. The bar is not CENTRALISED or not aligned with the axis of the lathe. Its wobbling to use the real technical term. So a 3 jaw is very convenient, but not accurate enough for critical work. If you can find your dimension within the bar as above, or its not critical - like a simple locking pin then fine. (there are specialist 3 jaws like the Griptru, and very expensive high precision ones, but the above is true for the normal common or garden 3 jaw that most of us use or can afford)
Collets - .probably centred repeatably to 1/2 thou?. Good 3 jaw .002-.003 TIR I guess when new, and a lot less if it has been overtightened or jobs grasped in the tips of the jaws. If it has to be something that is dead true, like a crankshaft or bearing housings in an axle tube, you can't turn it end for end in a 3 jaw and expect it to be straight enough to fit a set of reamed bearings and all to spin true. It won't. You have to set it true in a 4 jaw, or mount it betwen centres for that sort of work.
PM me and I'll give you a phone no if you need a bit of help - its a simple job, just a bit of a pain first time out..
Edited By meyrick griffith-jones on 30/09/2009 13:13:35
Edited By meyrick griffith-jones on 30/09/2009 13:16:23
Edited By meyrick griffith-jones on 30/09/2009 13:19:30
Edited By meyrick griffith-jones on 30/09/2009 13:21:28
Edited By meyrick griffith-jones on 30/09/2009 13:24:05
Edited By meyrick griffith-jones on 30/09/2009 13:28:04
|Peter G. Shaw||30/09/2009 15:06:58|
986 forum posts
Might I suggest that instead of using shims, people consider making jacking screws as per Tubal Cain in Chapter 9 (Lathe Alignment) of his book Workholding in the Lathe - WSP 15.
The idea is that these screws can be made regardless of how the lathe is set up, then once installed, adjustment becomes simplicity itself.
Peter G. Shaw
|1008 forum posts|
You probably could - but its not so simple on these bigger Chinamen because the mountings for the bolts go into slots on top of the pedestals/cabinets. On top of the plate in which the slot is located is a rubber pad to stop coolant going down the slot, so you'd have to make a plate, then lift the lathe off its mounts and then secure the plate and a gasket. . Then you can start jacking. But then you are supporting the weight of a very substantial lathe on 4 very thin bolts so there is then relatively little vibration damping.
So what you gain on the roundabouts, you lose on the swings! Its not actually that bad, because once you have done it, shims don't drift!
He's talking about a Warco 12/24. So he can get a couple of feet in between centres, and turn 17-18" diameter with the gap removed. Played the same trick getting a Colcheter into a friends place. Its just one of those things you have to do if you are getting into machinery of that size and power.
For Tubal Cain the standard modellers lathe was probalby a Myford or an Emco I don't know about an Emco, but a Myford 2 people can lift and position.(with red cheeks)
The Myford uses long bolts though mounting pads with built in O rings to seal:and there is a built in adjusting nut and a locking nut. You are quite right, it does make it very easy. These 6" lathes are rather bigger lumps and while you can use the same principle, the mechanics are touch more serious. Like you need a crane or hoist to get it installed and once installed or assembled dismounting it completely is certainly doable, but its not likely to be a popular option!
Of course the best thing would have been to have made sure the floor was perfectly level to engineering limits in the first place!
|Anthony Salisbury||30/09/2009 19:19:24|
|32 forum posts|
Thanks for the information I'll have a go with this. Do you have a drawing sketchof the collar test peice?
|1008 forum posts|
Not that critical Ant. Just a cotton reel shaped thing. As long as you have a couple of raised collars to skim on about what 8" apart and about 1/16" high and 1/42 wide is plenty. Sort of ruler measurements.
Metal dia about 1 1/2" - 11/4 on a biger lathe. Anything that's rigid enough not to spring under load. I have a bit of 1 1/8 ali HE30 from my racing days, so thats good enough, but if it has to be steel, it has to be steel.
Big heavy lathe - hold shims in long nosed pliers while you slide them under the foot - watch your fingers.
One point. The machine is secured to the floor? Cork tiles makes good cheap damping, and stops a lot of drumming in hte cabinets.
Its all a bit experimental - you have it bigger at the chuck end. Slacken off, pop a strip of .010 brass shim either side of the mounting bolt, say 3/4" wide, under that back foot there, tighten down firmly and see what difference it makes. It ain't that scientific!
Edited By meyrick griffith-jones on 30/09/2009 19:37:44
|Michael Gray||30/09/2009 23:38:49|
45 forum posts
One method which I found helpful, but perhaps not to everyone's fancy is:
Rollie's Dad's method of lathe alignment
found at: this site it's a pdf file, but not huge.
362 forum posts
0.1mm taper over 50mm seems allot even for a lathe fresh out the box, over the length of the bed it seems to me to be allot of twist to correct. I wonder if the headstock its self was properly set up in the factory. However, you would have to make sure the lathe bed is set up flat and straight before adjusting the head stock and their lies another problem, you are really relying upon this being accurately set at the factory. . I have the belt driven worco version and the head stock only appears to be held with 4 bolts and nothing to key in its alignment with the bed so it depends on it being properly set in the factory.
I also had a terrible time getting alignment. After a day of frustration I welded 2 substantial bars across the base of the cabinet (front to back) at each end and fitted 3/4" jacking screws. It still took me a couple of hours to get close; it turns perfectly parallel up to 50 mm from the chuck then tapers 0.05mm. This I have come to live with, for accurate work I either finish by hand, with emery, or try and keep the turning at a position where it turns parallel. As a result I have never felt happy with the machine as this problem takes away the pleasure some what.
This lathe has a gap bed, I have never removed the piece that fits in the gap, its my guess that as I adjusted the jack screw, thus twisting the bed, the piece in the gap does not twist with it and is now out of alignment with the main part of the bed, so I wonder if the head stock was ever correctly set up and should have been adjusted. I have decided not to loosen the head stock and make adjustments that way as I can see myself getting into a right old mess.
I work away from home allot and bought a Sherline so that I can make parts while away; I chose it because it is light and portable. What a delight to start using a machine with no issues, perfect alignment and real quality feel. This I am very happy with.
|Ian S C||01/10/2009 12:15:55|
7447 forum posts
|Hi Nigel,had the same problem with my 1324 belt head,I adjusted the head about 10yr ago,it now needs readjustment and I'll shim the bed.If I was mounting the lathe fresh on the stand(home built from 6x50mm angle iron with 75 x 50 channel pieces for the mounting points,on the tail stock end I would tap two holes outboard of the mounting holes andput in screws from underneath for adjustment.Might be a bit dificult with 275kg of lathe sitting on top!|
|Peter G. Shaw||01/10/2009 13:12:37|
986 forum posts
Fair comment. I hadn't realised the sheer size of the OP's lathe. It's just that my own lathe, Warco 220, is similar size to the Myford, and has four feet with a hole through. Hence Tubal Cain's jacking screws worked perfectly for me.
|1008 forum posts|
Peter - bit chunky compared with a Myford- as I found out when mine arrived!! Kind of out of my experience too when it came to finding space for it.
Traction engine 42. Are you sure the headstock is not aligned. I know that not all Chinamen are identical, but if you go to the back and open up the cabinet door, on mine plainly there is a machined right angle which locks up against the back of the lathe bed. Thats not to say that swarf or some other problem couldn't exist, but I think you'll find a proper regisert in there.
One can check on the 90 degrees if you are sure your faceplate is true. Take a DTI across the facepalte. You should pick up a small "error", because the lathe should be set to turn ever so slightly concave by a couple of thou per foot or less. But concave it should turn. So if the headstock deviates more than that, then you might start looking harder.
However this lathe hasn't been set up, so there is a good start. I might, if I was being thorough, take the chuck off the camlock face and ensure there there was nothing AT ALL (not even the smallest smidgen of a flake of brass - the common culprit) trapped between chuck and mounting face. And I'd also check the taper rollers were propery adjusted and locked up.
Also, if Ant don't mind, I suspect that a bit of unsupportive practise has been going on. If you habitually work with long overhangs then taper is inevitable, not from the lathe, but from spring in the work. (This is a nasty dirty habit which one ought to be out of a bit sharpish). But, an unaligned lathe bed combined with an unaligned tailstock can make things worse. So for me, first get the lathe straight, then align the tailstock, and then use the tailstock for its intended purpse - supporting any and all work where the overhang is longer than about 1.5"
The Rollies Dad system looks fine doesn't it?. I'm not sure it would be so accurate as a test bar in the morse taper, but I bet it would get you close, perhaps very close. It is after all the collar method without the cuts, and using an average error. At its very basic, it has to be a lot better than nothing.
Lathes cutting straight from new? Well to misquote Tubal Cain. A lathe bed is ground straight, then you start hanging offset weights like motors and gearboxes on a long slender body. Then you put it on a mounting platform locked to an uneven floor. Of course its going to be pissed, totally pissed, and the longer the lathe bed, and the heavier the lathe, the more flexible and the more pissed its likely to be. A little lathe like a Sherline or Cowells hasn't got enough weight to really flex anything. It's also very short so its relatively very rigid, so you can easily get away with it. A Myford is pretty assymetric, longer, and very slim so you won't. My 12/37 is so darn long I can flex it quite easily with a spanner on a bolt - one handed with no effort. With a laser aligner you can see the displacement in the bed as you jack it.
So they need setting up, and more setting up the bigger you go.
Ain't rocket science people.
Edited By meyrick griffith-jones on 01/10/2009 18:34:15
Edited By meyrick griffith-jones on 01/10/2009 18:36:03
|18 forum posts|| |
Sometimes the problem comes from the compound slide after hard work or a dig in. The axis of the compound slide have to be dead parallel to the headstock axis. An error of 0,12° is enough to produce a taper cutting like your’s on a 50mm travel.
Don’t trust the 0° setting of the angular dial of your compound slide.
To check it’s easy. It’s almost the same as checking general alignement. Loosen sligthly the nuts that lock the rotation of the compound slide. That’s to ease it for further light adjustment. You need to remove the tool post and place a lever indicator with a magnetic base on the compound slide. Then chuck a ground round bar of says ½’’ diameter. Wind the hand wheel to move the compound slide and place the indicator on to the bar as close as possible to the chuck.
Lock the longitudinal movement and the transverse movement (cross slide). Set the indicator dial at zero. Now move away at a maximum distance from the chuck with the crossslide handwheel. Note the reading on the indicator.
Rewind to go as close as possible to the chuck. Move the chuck by hand to find the maximum deviation of the indicator. Note it. Move away and note the reading
Then calculate A= (Near chuck1+Near chuck2)/2- (Awaychuck1-Awaychuck2)/2.
With a soft hammer and ligth blows -on the tail of the compound slide reduce or augment the reading on the indicator the reading by A/2. Then check again doing same procedure.
After two or three adjustements the total variation will be reduced to a minimum and won’t vary. Tighten the nuts. Check again
|1008 forum posts|
You are quite right - It could be the topslide.
I never use it while turning. - I set length to a shoulder on the saddle with a saddle stop, get the topslide so it doesn't overhang, and .005 short of target, and lock it. That gives me maximum rigidity and the least possibility of any chatter, or problems while parting off. Do all my turning with the topslide locked and then finally move it forward the 5 thou needed to get the shoulder length right, and then wind out to give a clean face. The topslide is set parallel with a DTI and never moves .
So with that method of working its not part of the equation.
On the Myford I have one of the Radford topslides from Hemingway. That has a tenon to set it parallel, and there is no taper turning facility - it doesn't rotate at all and it is guaranteed to be parallel. (If I want a taper I either use a home made taper turning attachment, or replace the original Myford topslide temporarily. the advantage of hte Radford is that the topslide never fouls the tailstock - one of the banes of Myford ownership if the topslide is set straight.)
So apologies for not mentioning it .
But, with respect, I'm not so keen on your system of aligning the topslide because you cannot guarantee that anything held in a 3 jaw will be true. His Warco has a topslide with a flat ground face that should be parallel to the topslide dovetails. So all he has to do is put a magnetic base on the lathe bed, and the foot of the clock on the side of the topslide and wind the whole saddle back and forth. Adjust the topslide till there is no indicator movement. After that one can use a setsquare agin the topslide to set the toolpost for say parting off - if youhave a quick release toolholder block to square up on.
If you are doing it on a Myford (or any other lathe with no flat face on the topslide), hold the clock in a toolholder (or remove toolpost and use magnetic base on the top), DTI foot against the side of the bed. Move topslide forwards and back. Rotate topslide till there is zero movement and lock down.
Edited By meyrick griffith-jones on 01/10/2009 23:09:53
|Anthony Salisbury||04/10/2009 11:16:17|
|32 forum posts|
After messing with my machine for months I followed the few simple steps and now my Lathe is running better than it ever has!!! Got it running 0.01mm / 6" or so. I'm really happy now feel as though I can do some proper lathe work without having to mess about all the time.
Thanks again the help is really appreciated!! I owe you one!
|11 forum posts|
You all may be interested in my experience with a 12/36 Tiawanese geared head lathe.
As received new it chattered badly but I persevered and after a few years had to replace the head bearings After ressambly the lathe cut badly tapered out of the chuck and I had to realign the headstock to correct the taper this process inproved the chatter markedly.
On examining the old bearing cups (tapered roller) I noticed uneven wear ie. wear in two diagonally opposite parts of the bearing track, the conclusion I reached was that the bearings had not been seated properly in the housing and upon installing the new ones, the resulting realignment moved the spindle and so caused a taper.
Incidently the correct way to align the headstock of a lathe of heavy construction is to check that the bed is not twisted and then move the headstock not twist the bed. Most quality lathes have provision for this by way of adjusting screws or at least oversize mounting holes.
Bed twisting is only used if the head cannot be adjusted as on light lathes like Myford or Southbend
Please login to post a reply.
Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!
You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.
Click THIS LINK for full contact details.
For subscription issues please see THIS LINK.