Accuracy of an old machine

hi,

I have a rather sturdy 1960s lathe which I'm in the process of refurbishing. My question is, it's seen a few shifts, what level of accuracy is good enough for general machinhing, say 2 tho over 6-8-12 inches?i get that in an ideal world parallel is the answer; however I don't think that's attainable without some expensive work.

Think about what you will be making, for the most part say 80% of my work a couple of thou is good enough. I make a fair amount of stuff for old motorcycles and lawn mowers and only chase a thou or less if the part requires it say on a bearing fit. A little more care and the job is usually OK.

The limiting factor is often the machinist, not the lathe!

I would prioritise getting a good finish when turning and being able to part off without difficulty.

You may well find that once the spindle is set up to work well it won't be far off turning parallel.

Neil

If you're not paying for your own time, then returning a machine to original specifications is not expensive. So far I've done a 1952 Hardinge HLV and a 19?? Beaver turret mill. 1976 Surface grinder will be done after I've re-lapped the surface table...

I've just turned down a 2" piece en3b, it's showing around 1.5 -2 thou variation over 6" the variation seems to kick in around 3" from the chuck, bed wear I suspect. The material was turned unsupported an thoughts guys? Further to marks comments, How much would It cost to grind the bed and where would undertake such a task? Thanks for the advice guys.

2 thou over 6 " unsupported is not to bad. On something a bit fussy you could most likely get that as good as spot on with a tailstock centre tweak the tailstock offset to correct error. Bed grinding place in the Nuneaton/Coventry area, most likely cost more than the lathe is worth. Bed grinding is just the start, you then have to refit the saddle and tailstock not to mention getting the workhead set up correct again.

.

My thoughts are that if that was me I would be a happy camper.!

Yes of course you could get it better but be aware that chasing tiny precision is a bath tub curve. - You can spend an awful lot of money for that little last bit of return.

I have been doing this engineering thing for nearly 3 years now so I am still very much a newbie. The machines I use are hobbyist ones but are far more accurate than I will ever be.

Unless you have very deep pockets you will need to learn the faults and limits particular to 'your' machine and learn to work around them or you will end up in a tail chasing frustration.

Nick

Any clue as to what make and model lathe you have? Makes a difference to how easy or worthwhile it might be to fix it up.

Tapered turning with no tailstock centre is not always down to worn bed. Headstock bearings are a common culprit and should be fully sorted before any bed work is carried out. If its a Myford, for example, the bearings can be tightened up with the removal of a couple of simple shims, or a little bit of judicious scraping if need be. Headstock alignment can be adjusted with two screws on the headstock and either some shims or scraping on the mating surfaces.

Also, "levelling" the bed, or more accurately twisting it, can be used on some lathes, eg Myford, to get it to turn parallel. This can be used as a dodge to compensate for worn bed.

Also on the earlier pre-1970s or so Myford you can do the "wide guide" conversion to abandon using the front vertical shear as the guide and put a strip of guage plate in the rear of the carriage to bear on the rear vertical shear of the rear bed way which is unused and still straight. This conversion has been covered in ME and MEW many times over the years, IIRC most recently in a recent (last few months) MEW article on refurbing a Myford.

Another cause can be bell mouthed chuck jaws, which can be reground with a toolpost mounted Dremel tool and the correct "cloverleaf" packing ring between the jaws.

Bed regrinding is a last resort, not starting point, for getting old machines to turn true.

Edited By Hopper on 24/04/2017 01:24:20

I'll try to shorten this as much as possible but..................

What type of tool was used to turn that 6" long piece down. To test like that you need to start with ultra sharp HSS since carbide takes a lot more pressure to make it cut unless it's one of the specialty tips meant for aluminum and has lapped edges. That pressure can change your results because it pushes the bar away instead of cleanly cutting it.With that very sharp HSS tool you then take a clean up cut with tailstock support. Remove the TS support then take a very light cut with that same very sharp HSS tool. .001" - .002" depth of cut. The general rule for cutting without tailstock support is 3-1 so your 2" diameter x 6" long just fits the rule.

Another test is a long heavy bar 2" + in diameter and again a light HSS cut along the whole bar until it cleans up and using tailstock support. Carefully measure that with an accurate and trusted micrometer at the tailstock, middle, and headstock end. More places is even better. Both these tests will give you some starting numbers. And the tailstock may well be slightly off center from the headstocks C/L. But you'll have the start of a good test bar.

Bearing checks and adjustments as mentioned by someone else should be done. Once your confident there well adjusted and in good condition then the bed has to be checked for twist. Getting it leveled with a high precision machinists level is the easiest method. The actual level isn't really important but you need both ways in the same plane with no twist. Given the ratio between what's happening at the bed mounting feet and the tool tip .001" at the bed foot can make .003" or far more at the tool tip. Actual bed wear has a lot less effect. Visualise your tool tip moving up or down .003" - .004" on a large 2" diameter bar. The movement difference will create very little change to the cutting tips location on the material or a difference in the cutting size. A .004" thou up or down movement does not decrease or increase the finished diameter by that .004". Bed twist causes the tool tip to move in or out and that does change the cut diameter by the same amount. It rolls the tool tip in or away from the material depending on which way the bed is twisted.With the bed level then the tailstocks morse taper can be checked by swinging an indicator to the front and rear just inside the morse tapers bore. This DOES NOT work for checking the tailstocks alightment for up and down. Gravity can and will throw those results off. As much as .010" or more. Once the tailstock is aligned and with all of the above done I'd then take another light cut on the long bar with a good dead center and tailstock support.

The measurements taken then will have some meaning about what your lathe is capable of doing. The quick test you've already done isn't of much use because there's too many variables that can sway the measurements. If that cutting test on the long heavy bar with support shows some real improvement then it's time to repeat the test you've just done.

Flat bed lathes like most of the Myfords and some others are the cheapest to have reground. But it's not quite that simple. Doing so lowers the carriage by the amount that was ground off. That affects the half nuts position on the leadscrew. And the same if you have a seperate rod for power feeds. If the bed is worn enough to need that regrinding the bottom surfaces of the carriage will need it as well. So more material is then taken off. To regain the correct alignments material such as moglice is bonded to the bottom of the carriage to lift it back to the original position. The mounts for the leadscrew and possibly gearbox if equipped with one could be lowered to regain the alignment, but adding the moglice is much easier. There's a bit more than all this, but it should give you an idea of what's roughly involved and getting some numbers that can be trusted.

There seems to be more misinformation and lack of understanding about testing and aligning machines tools on every forum I can think of than there is correct information. I've also seen multiple posts on just about every forum about just jumping in and moving the headstocks position. That's the very last thing you do without properly tested and verified results that do show there's no doubt it's actually out of alignment.Googleing for the online PDF of what Dr. Georg Schlessinger wrote about machine tool testing is well worth the time to read it enough so it's understood what's being done and why. I've had to spend a great deal of time, research and hands on trial and error trying to learn the very little I do know about it. There is more than a few ways of doing this with some more complicated methods than others. But I know what I've mentioned does work.

Hi all,

Some welcome thoughts, its a Lang j6 13" swing. I supect the bearing need adjustment too, I've placed a bar in the chuck and lifted the chuck with a piece of 2x4, there seems to be about .003 play. The bearings have shims in that have to be ground to remove the play. The problem Is i don't have a surface grinder. Also at some point the tool post has been run into the headstock, I'm therefore a little concerned it's shifted the gear head; however I'm a little loathed to start messing with that without concrete evidence it needs adjustment.

I've borrowed a test bar a few weeks ago; however the accuracy of it seemed a little jubious. My next action was going to be to buy a new test bar with a morse taper and stick it in the spindle. This leads on to my next problem, I don't have a reduction sleeve, in the manual it talks of it being a morse taper reducing to mt3; I've tried a 5-3 and it's too small the next size up will be too big. Any advice on where to buy one? I could try to make one on said machine I guess.

I'm not familure with a Lang lathe, but they all work much the same. .003" seems high though. For the adjustable type preload you slowly tighten, run at top speed for 20 min. to 1/2 an hr and check the headstock temperatures with your hand. Keep adjusting until the headstock starts to warm.Proper preload is very close to being correct at quite warm but not so hot you can't leave your hand on it for some time. A bit subjective but that's what most of the recomendations I've read have to say. You might find ground shims at a good bearing supplier and add, subtract in various combinations until it's correct.

Likely that's a 4 1/2 to 3 morse taper. Fairly common on a lot of lathes in the 5 MT size and not usually too tough to find by contacting a few machine tool suppliers. Memory if it's right then the Colchester Chipmaster used it as well. It usually takes a fairly serious crash to move a headstock on a well built lathe, but possibly that was enough.

Ditto, not familair with a Lang lathe. Some pics would be nice, so we can see what we are talking about.

That said, yes .003" movement at the headstock spindle is at least .0025" too much. More like about .003" too much. There is absolutely no point at all in proceeding any further with alignment etc until you have fixed the headstock bearings. This amount of slack would be enough to create the taper on your test turning piece. So you may not have a bed wear or bed alignment problem at all, just a headstock bearing adjustment problem.

Depends on what the shims are like, but usually you can cut out new shims from shim stock and insert them in place of the old ones. How thick are we talking about? If they are big enough, you might be able to very carefully file them down, using a mike to make sure they come down parallel. Or you may be able to rub them on emery cloth on a sheet of glass to knock a few thou off. It does not take much to remove the .003" you seem to require.

EDIT: PS just had a squiz at Lang Junior on the lathes.co.uk site. It looks like if you have the roller bearings, there would be a large circular shim that goes under the collar that provides bearing pre-load? You should be able to cut a new one out of shim stock that is a bit thinner than the original and try it out?

The website also indicates you have a very good quality machine there "better than Dean Smith and Grace and many German machines" is pretty high praise. So well worth persevering with, cautiously so as to not overdo things.

I'm not a big fan of spending money on ground test bars. You can turn one up yourself between centres if you sort out that headstock bearings and then the tailstock alignment. But even then, final adjustment is best done by turnin a 6" long piece of bar as you have done. This gives you the "real world" reading of how the lathe is cutting, which takes into account wear in the carriage and cross slide etc and the forces on them under cutting load. Static alignment is one thing, working alignment another.

It seems unlikely that a carriage crashing into the headstock would budge it on a solid machine like this. I woiuld not worry about that possibility too much.

You need to measure the length and the diameter at each end of the headstock spindle taper. It may not be Morse, but some odball like Brown and Sharpe or Jarno or a couple of others that used to get used. Or it could be a Morse 5-1/2 which is a MT5 but only half the usual length.

Edited By Hopper on 24/04/2017 09:23:06

Look on Tony's     Lathes.co.uk site       http://www.lathes.co.uk/lang-junior/

 

lots of info

Edited By KWIL on 24/04/2017 10:35:46

An impressive and extremely heavy duty lathe for there size. Anythings possible of course, but after seeing those pictures from KWIL's link (thank you) I'd be even less concerned about that crash shifting the headstock just as Hopper said. And I think much the same as he does about spending the money on those test bars. First they have to be made very well with as close to zero runout as possible. Getting that accuracy no matter where there made costs a great deal of money.Then the spindle taper has to be in perfect condition or the tests don't mean much. I have some and have learned that for what they cost the return wasn't worth it. They are handy for checking and aligning a tailstock not so much for a true C/L with the headstock, but for the tailstock being angled off side to side or pointing up or down hill. But even that can be done without having one.

Your Lang lathe certainly has the quality to justify a proper rebuild. If you ever decide to do so it's best not to chose to have the grinding done by price alone. The shops that do know what it takes to correctly align and regrind a lathe bed will charge accordingly. As Hopper said sorting out your headstock bearing clearance should be the next step. With that .003" clearance it's impossible to judge how much the bed wear is causing a taper in the workpiece and how much the excessive bearing clearance is. Logic and some guessing suggests it could be causing most of that .002" taper. It is a very heavily built lathe, but even so it's not impossible for the bed to twist and on an uneven floor, even more so if it's been bolted down without being sure it's level or not twisted in all directions. That could have happened at any time before you bought it. The lathe carriage then wears to the beds position and the bed can also take on a permanent twist over time. If so it needs to be shimmed and gently forced back into to being true.

To test for play in the bearings it states the following in the below photo, any idea where I can get a 200lb spring

No photo visible. There is a drawn out process to post pics on this site. There are instructions somewhere but not sure where. Basically put pic on Photobucket or similar and copy the DIRECT link and then paste it in the dialogue box on your post on this site that pops us when you click on the tiny square icon immediately to the left of the camera icon at the top of the post pro-forma.

If they are saying use a 200lb spring etc, a good alternative might be a spring scale used by fishermen or one made to weigh your luggage before departing to the airport. Or even a bathroom scale may be jury rigged to provide 200lb of force in the right direction, providing the domestic authorities are out shopping etc.

To post a photo, you need to go to the top green line, and click on Albums. Reduce your photo to about 640 X 480, and use jpeg image. you then can transfer the image onto your thread by clicking on the little camera in the top row of the message box.

Ian S C

All,

Thanks for the advice thus far, in response to one of the comments, I have the manual from lathes.co.uk. I've done a little more work, unsupported the machine cuts parallel for around 75mm but the suddenly cuts a .05mm taper towards the headstock over the remaining 75mm. (I.e. The dia of the work piece reduces) I did wonder about bed wear but surely it would do the opposite? It has been mentioned that the headstock could be out of alignment, I'm a little sceptical my self. I've put a new gasket between the saddle and the carriage, I did wonder if that could be causing issues. Any thoughts ??

Matt

The .003" of play you said you have in the headstock bearings would cause it to turn smaller toward the chuck, larger toward the tailstock end of the job. You need to get rid of that play in the headstock spindle bearings before you can make any meaningful observations about what might be happening with the bed and carriage side of things.