Lathe turned surface finnish

[maninshed]

Has anyone got any experience with turned surface finish on lathes. I've been through all the usual, tool grinding, tool finish, vibration, alignment etc etc. What I seem to get, is rings around the turned bar at regular inervals every 15 to 20 mm, almost as if the feed speed has been interupted? Wondering if its a gearbox problem or feed gears. From a recent thread I did make a tangental tool holder similar to the one in Tels photo's. This improved the finish but the rings still are there. Any thoughts??
The lathes a 5" with norton gearbox, feed driven through keyed slot in leadscrew.

[weldsol]

Question is it better with or without coolant ?

Paul

[brozier]

Another couple of questions.

Do you get the same effect when hand feeding? If yes could be an issue with bed not being completely flat.

Is the motor single phase? Single phase motors vibrate a lot more than 3-phase. I think this can be a cause of this sort of issue. I have a theory it sets up a resonance in the gear train but I coul dbe wrong!

Cheers
Bryan

[maninshed]

Paul, it makes no differance if I use coolant or not, if anything the finish (final cuts) is better without coolant.
Brian, no I dont seem to get the ringing so much with hand feeding, that's one of the reasons why I was wondering if if was a gearing issue. The lathe was fitted originaly with a single phase motor, now it has a 3 phase one run from an inverter, changing the motor did'nt change much. Was thinking about your thoughts on resonance in the gear train, I wonder if I substituted one of the gears in the gear train from the headstock for a plastic gear. Thinking about it quite alot of machinary manufacturers fitted tufnol or nylon (type) gears in the headstock train, think Myford do and Harrison used to. I always thought they acted as a weak point, (something to bust if things went wrong) but having a change in material would change the resonance as you suggest, you could be on to something.
Thanks lads for taking the trouble to reply
Regards Martyn

[weldsol]

You say the feed is through a norton box if you change the feed rate does it make it better or worse ?

Paul

[alanstepney]

I recall seeing somewhere, (may even have been here) about similar problems with surface finish.
In that case it turned out to be the main drive belt, and changing it cured the problem.
From memory, the lathe concerned was a ML7/Super7.

Might give a clue of a direction to look.

[kneedeepinswarf]

In my experience, poor finish that is due to vibration usually shows as a continuous pattern along the length of the work-piece, not the solitary ring at intervals that you describe.
I think your problem could be caused by a minute particle of work-piece material becoming welded to the tool tip, causing the tool to cut deeper, and then breaking off, leaving the ring. This can often be the result of too fine a feed. Popular opinion seems to be that a fine feed is needed for a good finish, but this is not always true.
I find that a sharp tool with the tip rounded to a radius of about 20 thou, well polished with a Arkansas stone, run at moderate speed and fed at 5 to 10 thou per rev, or even more, with a supply of straight cutting oil (not lubricating oil) will give good results.

Lionel

[houstonceng]

I had the sort of finish Lionel was talking about in his first sentence, a spiral of "chatter marks" on finishing cuts with auto-feed. Original single phase motor was removed from the lathe (rigid mount on headstock) and mounted on the wall behind to test out the theory. No marking.

I had, of course, tried all the other variables like tools, speed, feed, material, belts, etc and then found in a very early MEW that someone in Oz with a similar lathe had the same problem with motor vibration.

Investment in a 3-phase motor/Inverter combo solved the problem with the motor now back on the lathe.

[brozier]

Alan has a good point,

Myford now recommend a segmented belt for their ML7 and Super 7 lathes. Both Motor to countershaft and countershaft to headstock. I have bought one for my lathe just haven't got round to fitting it yet!

This has two advantages - you don't have to strip down the headstock to replace it and vibration is reduced.

I remember one of the Myford fitters waxing lyrical about the improvement on Single phase lathes.

Martyn has your drive belt been left tensioned in one position for a long time or is it just old? might be the cause of some of your problems.

Cheers
Bryan

[russell]

I wholeheartedly support the idea proposed by several posters that replacing the single phase motor with a three phase and inverter improves the surface finish. I have just done this as part of a rebuild of my Atlas and the improvement is well worthwhile. With care I can now get a near mirror finish.

However, as Lionel said, this is not Martyn's problem. In the past I have noticed rings in the finish at 1/8 in. intervals (lead screw pitch). This was cured by slightly looser adjustment of the saddle gib. It was possibly caused by the longtitudinal slot (for crossfeed) in the leadscrew.

But, 15 to 20mm period? Is it regular or does the pitch vary between 15 and 20? That may give some clue.

Regards,
Russell.

[maninshed]

Hi guys, lots of new ideas to try here and thanks to you all for the input. The belts are coventional and are original from when the lathe was new 1989, have thought about trying the Nulink belts to reduce any possible vibration in the headstock, but finish looks okay when hand feeding. Will try the idea of a lightly courser feed never really thought of that one. Do'nt think it's the leadscrew as the feeds drive through a keyway in the leadscrew.
Russel mirror finish i'm jealous, maybe if I work through the advice above I'll hit the jackpot, thanks again for all the support.

[steammadman]

3 esentials 1 very slow speed
2 very fine feed
3 very very sharp tool (HONED WITH A GOOD QUALITY OIL STONE)

That should work for you oh and NO MORE THAN 3 thou cut

[havoc]

1 very slow speed

My experience is that this depends on material and tool. When cutting C45 with a TC bit a high speed gives better finish then when cutting brass with a HSS bit.

[drjohn]

Hi mainshed - I have posted this picture long ago and asked about the finish - I have experimented considerably with my cheap Chinese lathe with single phase motor attached to the lathe and lots of vibration transmitted everywhere, but the answer seems to be tool tip form and surface speed.

From the picture you can see the previous cut was as you described, then with nothing changed except the O.D. of the workpiece by the present cut, the finished changed dramatically. You can also see that at the end of the previous cut, the finish changed - maybe to do with minor wear on the carbide insert? dunno!



The workpiece is just under 1" diameter mild steel, turning at 375 RPM and the cut is 0.5mm with a feed rate of 0.1mm (mixing my metaphors ) and as you can see, no coolant.

[jonkett]

With a decent grade of carbide you should be at about

Diameter of work = Dm
surface speed = Vc 200 - 300 m/min
feed rate = Fn 0.2 - 0.5 mm per rev
depth of cut > 0.15 mm
RPM = n

Calculation is :

n=(Vc*1000)/(.3417*Dm)

n = 2546

[drjohn]

Oh! ? I got it seriously wrong but it worked - That's model engineering.

Flat-out my lathe only does 1360 RPM and at that speed I have to chase it round the shed as it pulls the bolts out of the wall and floor with the vibration.

[Deleted]

Hi DJ. I'm nor waffling this time 'cos this is my subject.
Jonkett is spot on with the revs for an industrial machine and a reasonable tool life. However, I would at least double your spindle speed and the feed to obtain consistent results. Optimum depth/feed ratio is 10:1 for our type of work, so 1mm depth and 0.1 mm feed would be ok for HSS tools, but double the feed for carbides. The work should remain quite cold as the heat should go into the chip.

In addition, just to complicate matters, your situation is not helped by:
1. A high tool post (they rock and wobble, especially when using a QC toolpost)
2. A tool that looks more like a threading insert than a plain turning type (again, rigidity is affected)
3. What looks to be an extended live centre, permitting the work to deflect unnecessarily (again, rigidity is affected)

All in all, you might have just been lucky to achieve that finish....
Regards
JB

[jonkett]

ok so no one spotted my deliberate mistake, .3417 should of course be pi. No idea what i was thinking of there.

The point i am making is that, yes they are figures for an industrial setting, but carbide (the photograph showed a coated carbide insert) does need to be made to work for the best results. Carbide hates a rubbing action which kills it faster than anything.

The other thing that was not discussed was choice of carbide grade and chipbreaker style, two elements that have more effect than anything else but that is a complete science on it's own.

If you can obtain Sandvik's Corokey book, should be available from any Tool Merchant or direct from Sandvik it shews all sorts of various set-ups and gives 3 options from perfect rigidity to what they consider to be the least optimum. Be aware that it is promotional item and is only dealing with the very latest technology but it does give all the basics and remember that they still produce many older grades that are more than adequate for our use.

Have you tried Cermet grades for finishing? I have been machining a Cast iron back plate, they are also used for finishing steels, and very pleased by the results. That is also on a Chinese lathe that would be better bolted to a concrete floor.

Incidentally tool life from all major manufacturers is defined as 12 minutes cutting time per edge.

[alanstepney]

Interesting to see the comments on carbide tooling.
Although I have used superb industrial machines (DS&G, Colchester etc) my lathe is an ML7, and the rigidity and speed mean that I stick to HSS, or even carbon steel, tools, which suffice for the vast majority of work.

I suspect that I'm not alone in that.

I find the "secret" of a decent finish is to get a REALLY sharp tool, honing the edge after grinding being essential.
Then, speed and feed to suit, being a mix of experience and trial & error.

[havoc]

Jonkett, I assume that Dm is in mm in your formula, otherwise I arrive at rather funny speeds (1270000 rpm for a 50mm diameter). With the diameter in mm I arrive at 1200 rpm which is more sensible.

I turned my last batch of 25mm wheels at 1000 rpm. The chips that come off then are a nice deep purple and very hot. More rpm I didn't dared to use.