By continuing to use this site, you agree to our use of cookies. Find out more
Forum sponsored by:
Forum sponsored by Allendale Jan 24th

Lathe alignment. What is good enough?

Late RDM Alignment

All Topics | Latest Posts

Search for:  in Thread Title in  
Skarven05/11/2011 20:24:11
93 forum posts
11 photos
I have had the 330x1000 Lathe for 6 months, and I wanted to check how accurate it was. Running Rollie's Dad's Method test I found that the headstock spindle was out of parallel with the ways 0.05mm for a distance of 350 mm. (This was after leveling, before that it was 0.17mm) The Headstock have adjustment screws, and after some serious consideration, I tried these, finally ending up with an error of 0.005mm for 350 mm. Getting below 0.005 was difficult, as just pushing a little on some part of the lathe was enough to to change the error by more than that.
This test was run with a 12.7mm (1/2") 450mm bar from an old printer chucked in the three jaw chuck. Later on I used a ER32 collet chuck giving the same result..

The vertical seemed to give some strange result which I think is because this
12.7mm rod is sagging. An aluminium tube 25mm OD 5mm wall thickness gave the same error horizontally, but no (0.01) vertical error.
How little error is good enough?
Steve Garnett05/11/2011 21:38:11
837 forum posts
27 photos
Posted by Skarven on 05/11/2011 20:24:11: The Headstock have adjustment screws, and after some serious consideration, I tried these, finally ending up with an error of 0.005mm for 350 mm. Getting below 0.005 was difficult, as just pushing a little on some part of the lathe was enough to to change the error by more than that.
How little error is good enough?
I'd say that having errors at this level over 350mm was pretty damn good! The moment that pushing parts of the lathe introduces greater errors, you've reached about as far as you can go, I'd say. And whilst printer carrier bars are generally quite accurately ground, I doubt whether they are much more accurate than you are measuring anyway. I mean, you are talking about a handful of microns here! But you could be right about the droop - my Arrand bar is somewhat thicker than that, presumably to eliminate this sort of error.
Skarven05/11/2011 22:16:21
93 forum posts
11 photos
I also feel that this is the limit of accuracy that I will get.
I will do a turning test tomorrow to check if practice matches theory
Of course, I still have the alignment of the ways in the gap to consider, and this has no horizontal adjustment. All the above measurements were made on the ways to the right of the gap. Does anybody have any suggestions for this alignment?
ady05/11/2011 22:19:58
612 forum posts
50 photos
Even if the error is zero once you start shoving toolbits into metal workpieces you start to introduce errors via various forces.
So the easiest way to do things is to have it acceptable for roughing out then get it as good as you can for the final cuts...and only experience, nothing else, will get you to that high level of accuracy we all crave.
I've found DRO great for general work, much easier than dials.

Edited By ady on 05/11/2011 22:22:16

blowlamp05/11/2011 22:44:08
1298 forum posts
83 photos
There was a programme shown recently, which followed the construction of the Rolls Royce Trent engine. One engineer was proudly explaining that they allowed a tolerance of 0.007mm (7 microns) for the grinding of the base of the turbine blades.
So with a result like you've got, I think you (and Rollie's dad) should apply for a job there
Once you're at that level of parallelism, you'll probably find that errors in parallel will now jump from end to end of the workpiece, depending on cutting conditions and the type of tool in use at the time.
ady05/11/2011 22:54:19
612 forum posts
50 photos
Some jobs take days because moderate heat introduces errors.
A finished job which is spot on for cold fitting will alter if it has any heat when it has to do any work.
Hugh Gilhespie06/11/2011 18:23:14
130 forum posts
45 photos
Hi Skarven,
Can you give a link to the 'Rollie's Dad's method that you used for checking alignment? I have looked at a couple of sites that give this method and they seem to contradict each other so it would be interesting to see which one you used.
I am just about to try and set up my lathe using this technique - or I will be when I have finished making new six new adjusting feet - why Colchester think 6 feet are required is a mystery but they do, so 6 off feet it is.
I have an Arrand test bar all ready to go and a ludicrously accurate and cheap dial indicator from CTC Tools, the dial reads in microns and it has a 1 mm range. It does seem to be accurate as well, or at least it agrees with the DRO on the lathe. Quite frightening to use as you think the run out is huge until you realise that it's only a quarter of a thou that's making the neeedle move 5 or 6 divisions.
Just need to get the beast level.....
Regards, Hugh
Steve Garnett06/11/2011 18:52:50
837 forum posts
27 photos
The most succinct explanation of the method I've found is this one. Won't tell you anything directly about the accuracy of the ways, though, or anything else about the saddle. And as far as I can tell it relies entirely upon the bar being very accurately round, even if the diameters don't particularly matter. Or at least the errors have to be in the same positions on the circumference at the measurement points. And of course it actually relies on 'incorrect' mounting of the bar to work at all!
I'm sticking with the Arrand, myself...

Edited By Steve Garnett on 06/11/2011 18:54:53

Skarven06/11/2011 19:09:07
93 forum posts
11 photos
Hi Steve
I agree with you. If I had a test bar, I would use it, but I don't.
Are they very expensive?
Hi Hugh,
My link is the same as Steve's
I found it at this location witch have some more alignment stuff.
I am doing some testing with a roundstock 50mm dia and length 800mm.
The lathe has been running the whole day with very fine feeds and cuts.
One trip along the rod is about an hour! Unfortunately it had a rough surface, so i have only just got down to a "round" stock. About one kg of swarf.
Maybe this could be the axle for a 1:1 scale narrowgauge model!
Hugh Gilhespie06/11/2011 22:07:44
130 forum posts
45 photos
Hi Steve and Skarven,
Thanks for the link. I do have that one - and some others that disagree with it! But, I guess it works so I will stick with it.
What I am planning on doing is using my test bar as the rod in the Rolley's Dad's method. It's a bit shorter than recommended, about 6 inches after the taper but at least I know it is accurately concentric.
Yes, they are a bit pricey, about £80 for the MT3 as I recall, but they are beautifully made and I decided to invest in some kit that will let me set up the lathe as accurately as I can. That way, I will know for certain that all the cock-ups are down to me alone! I have made a couple of dedicated DTI holders that fit into the normal tool holders on the QCTP, one for horizontal and one for vertical alignment.
The idea is that by having the kit easily to hand, I can check and adjust the lathe on a fairly regular basis. I also intend to make a face plate attachment for mounting a DTI to allow accurate re-centering of the tail stock - knowing that I can re-centre it easily I will be more inclined to use a set over tail stock for taper turning, etc.
I have also paid a bit of attention to the adjustable feet for the lathe. These need to lift the lathe about 4 inches so I can use a pallet truck for moving it as required. However, I didn't fancy long studs, too much vibration, so I am using lengths of 2 inch bar for the main supports with fairly short studs that will allow plus or minus 0.5 inch height adjustment. The studs are M12 by 1 so one full turn of the nut will give a 40 thou height change. The nuts will also be made from 2 inch bar and fitted with tommy bar holes. Conventional hex nuts will be used to lock the main nuts after adjustment. I will use soft copper washers between the nuts and the lathe to bed down onto the casting.
If I can get anywhere near the accuracy that Skarven has achieved I will be a very happy bunny.
Regards, Hugh
Steve Garnett06/11/2011 22:51:32
837 forum posts
27 photos
Arc Euro Trade do a cheaper MT3 test bar for £50. We have one at work, and I've compared the relative accuracy of my (roughly £90) Arrand, and this, and I have to say that you get what you pay for. The Arc one isn't bad, but it's not a patch on the Arrand.
Yes the Arrand is a bit on the short side, but even over the length you get you can make very accurate measurements with some confidence. And if you can get your tailstock accurately aligned, you can extend the measurement range by placing it between centres (assuming you have a good headstock centre, that is) and reversing it.
If you use the Arrand for Rollie's Dad's method you may run into a few problems, especially if your headstock is in good condition. If it's locked into a clean taper, then essentially you won't be able to: a) spot any difference in the diameter from one end to the other, and b) be able to measure any eccentricity - it's too damned good! What you will find is that with a dial gauge on the carriage aligned exactly at 90 degrees to the vertical, you'll just detect the combined error of any misalignment of the the headstock, and wear in the ways - as a direct measurement. Chances are that if you are running on a V bed (ie not a Myford), then it will be the headstock misalignment that constitutes most of what you've measured. Then you have the whole rigmarole of working out where the real error is, what's caused it and whether you can do anything about it.
Every time I think about this, I imagine the nightmare scenario where the headstock is misaligned, this is mistaken for bed twist and corrected appropriately. And that's the snag with Rollie's Dad's method - this really could happen, because you have only averaged diameter references - no absolute. When it comes to the crunch, I think that the test bar is your only real path to salvation.
blowlamp06/11/2011 23:36:41
1298 forum posts
83 photos
The trouble with a test bar is that it can only be as good a reference as the taper into which it is mounted.
Aligning a lathe headstock is straightforward once it has been established that the bed is not in a twisted state (by using a precision level) and can checked and adjusted by gripping a bar in the 3 jaw chuck and relieving it in the middle, so as to leave two collars that can be repeatedly skimmed with a sharp tool and their diameters compared, until within your chosen limit.
For me, doing it this way removes the confusion from have a perfectly parallel test bar that is actually running off centre because of a slightly inaccurate taper in the spindle and also yields a 'real world' result.
Steve Garnett07/11/2011 10:00:01
837 forum posts
27 photos
Posted by blowlamp on 06/11/2011 23:36:41:
The trouble with a test bar is that it can only be as good a reference as the taper into which it is mounted.

With the Kerry I have, I've checked the headstock taper for concentricity as far in as I can, and it's pretty much spot-on. In fact it's so good that I thought the dial gauge had stuck; I think it's had very little, if any, use during its previous existence.
But the Arrand bar has uses beyond that - you can mount it between centres as well.
In a practical sense, you're right though - you want to check how it goes with turning, as that's the whole point really!
Terryd07/11/2011 14:22:29
1933 forum posts
179 photos
Hi All.
I don't know what you guys are making but your projects must be amazing. We never had equipment this good even in our inspection dept. I am reminded of the following quote from the writings of Martin Evans (editor of Model Engineer 1965 - 1977).

"If I did not know Henry Greenly personally, I knew one of his chief mechanics, A. P. Campbell, very well…………………….
Campbell was a brilliant craftsman of the old school. Give him a few hand tools and an old lathe of doubtful accuracy and (my italics) he could turn out almost anything. Readers may have seen samples of his locomotive work in two beautiful Great Western Railway models, a 'Castle' four-cylinder 4-6-0, which is in the Science Museum, South Kensington, London, and a Dean 'Single' 4-2-2 which is (at the time of writing this) in store, but formerly used to grace the station entrance to the Great Western Royal Hotel, Paddington."

Good luck with your projects

Hugh Gilhespie07/11/2011 17:05:51
130 forum posts
45 photos
I do have some issues with the headstock taper. With the test bar fitted, I get a TIR of about 0.5 thou close to the head and about 1-2 thou at the end of the bar. The actual taper in the spindle is an MT5 I think and I have an MT5 / MT3 adapter that came with the lathe - a Colchester Student 1800 of 1975-6 vintage. The adapter has taken a few knocks in the last 35 years and is a bit battered. This is why I am looking at Rolley's Dad's method for initial alignment . I do agree with Martin that the final and most useful test is by measuring diameters after careful cuts.
I also know that the bed is not level and I am hoping that getting it level and twist free will do everything I want. I will adjust the headstock if I have to but I am hoping fervently that I don't have to!
I think the spindle bearings are OK - or at least, when I clock on a freshly turned piece of bar there is no measurable run out - none detectable at all and I hope that means that the bearings are good.
Hugh Gilhespie07/11/2011 17:16:00
130 forum posts
45 photos
Hi Terry,
My projects ARE amazing. At least in the sense that I am amazed that I have actually made something that actually works! Doesn't matter if it doesn't work very well or all the mistakes along the way - just being able to create something myself gives me a real thrill. It's not false modesty when I say I am not very good at this, just the simple truth but you have to start somewhere.
Not only do I like the idea of having the basic kit in good working condition, for someone with as little experience as I have, it is an absolute necessity. I just don't have the craftsmanship or experience to fall back on so I rely totally on the equipment I use. If I am feeling optimistic, perhaps I will get better and develop more skills but it's not guaranteed and in fact not even needed. I get a huge amount of satisfaction from what I can do now.
Regards, Hugh
Skarven07/11/2011 20:04:09
93 forum posts
11 photos
Hi Terry, Hugh,
My project is also amazing. I'm trying to turn a 50x800mm rod to the highest accuracy I can get. I'm a beginner, and it's not that easy to to turn to a specific diameter with a close tolerance. I'm also trying out how to restart a turning if you have to stop it for some reason. That is difficult!
I also have problems with the carbide-tipped tools I use. It is not possible to take cuts of less than 0.02 (Dia 0.04) This will chatter and make a bad finish.
My result so far is to turn a length of about 600mm of to a diameter of 48.50 +/- 0.015, but that is to the right of the gap in the ways. I still have to do something about the gap (mind the gap
This is real fun!
Dusty07/11/2011 20:45:14
471 forum posts
8 photos
Tipped tools are not the best if you are taking small cuts. They work best if given something to do. It is also speed dependent carbide and the derivitives work best at much higher speeds, as you do not say what the material you are turning is it is difficult to advise but mild steel with a carbide tool I would suggest about 450 rpm for 50mm dia
My preference would be for HSS tool and a speed of about 250 rpm. As for stopping and restarting I assume a finishing cut, my advice is don't. Are you using the tailstock to support the end of the bar? If you are not you will never get a parallel workpiece and the surface finish will also suffer. Without the tailstock even 50mm bar will deflect, it will also try to climb over the tool.
And another thing before you get involved in Rollie's dad's method make sure the headstock bearings are properly adjusted, because if they are slack then everything else is to no avail.
I wish you the best of luck and remember on your finishing cut plenty of cutting oil/mystic/suds/soluble oil call it what you will, it is the only way you will achieve your goal over 800mm.
Steve Garnett07/11/2011 20:47:43
837 forum posts
27 photos
I'm amazed whenever I get anything to work properly!
Skarven08/11/2011 07:01:37
93 forum posts
11 photos
Hi Dusty
I know HSS is best for this kind of work. My problem is that I only have a very bad bench grinder and I,m no good at handheld grinding. I'm going to make a tool holder something like the one in MEW 179. "A quick and easy tangential tool holder".
I really don't know what material I have. It was just lying there in a scrapyard and looked like a nice thing to have. So I bought it for about 10£. It's a bit harder than mild steel.
Most of the turning has been at 160 rpm, cutting long threaded swarf, but I have had to reduce to 65 once because of chatter. Maybe an increase would have been better.
I also use oil. Actually, after a lot of testing with neat cutting oil, soluble oil, WD40 I tested Olive oil and rapeseed oil. The rapeseed seemed no worse than the neat cutting oil in my tests, cost next to nothing, and has the great advantage that it does not smell and give you all sorts of skin ailments in case of contact. It also cleans up real easy.
As for restarting finishing cut, this is only training, but some day maybe the tool wear out during a cut, or some other stupid mishap...
Yes, the turning was with the tailstock support, but I was able to turn another piece of this 50mm rod with a length of 200mm only supported in the 3-jaw chuck. Small cuts and the finest feed of 0.053mm/rev.
I know it seems a bit stupid to put so much work into something you are not going to use, but I really need the training, it's a lot of fun, not that much work because of the power feed.
I can also do some milling in between.

All Topics | Latest Posts

Please login to post a reply.

Magazine Locator

Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!

Find Model Engineer & Model Engineers' Workshop

Support Our Partners
Eccentric July 5 2018
Allendale Electronics
Eccentric Engineering
Subscription Offer

Latest "For Sale" Ads
Latest "Wanted" Ads
Get In Touch!

Do you want to contact the Model Engineer and Model Engineers' Workshop team?

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.

Digital Back Issues

Social Media online

'Like' us on Facebook
Follow us on Facebook

Follow us on Twitter
 Twitter Logo

Pin us on Pinterest