Finnish on Test Piece

I've managed to get my Myford Super7B turning with the aid of an inverter and have wired in a potentiometer and switches for on/off and rev/fwd and all is working. I have no precession level so have levelled the lathe the best I can with a cheap 6" bubble level and am turning a test piece of steel about 6" long and 1" diameter (relieved in the middle section) and can get the two rings pretty close to the same diameter, however the finish on the section close to the 3 jaw chuck is smooth like a mirror whilst the outer section is rougher and grooved and makes a chattering noise when I turn it. I have tried different lathe tools and tool angles with minimum overhang but the outer ring is always of a poor finish compared to the inner ring. Any thoughts?

Lack of tailstock support and/or play in the spindle bearings

Edited By JasonB on 26/05/2022 17:12:39

Unsupported length has probably hit a resonant frequency, needs support at outer end to eliminate the chatter. Dave W

As the OP is turning a test bar , I'm not sure a support is a good idea ? Mirror finish might indicate a rubbing tool or poor tool grinding, above centre height tool. OR the spindle bearings need adjustment ? Good Luck, Noel.

6:1 for length:diameter is poor, if there is no support from the tailstock - OP does not provide full details, so I would guess this is the case, and reason, for the different finish. If the surfaces finishes were ‘tidied up’ with emery, I doubt they are as close to each other as presumed.

google either Roly's or Rollies Dad's method. Turning a test bar stuck out that far without tailstock support always seemed like a bad idea to me, and you can't use tailstock if you're trying to eliminate bed twist.

A good start is to have a long bar in the chuck with a clock against it at the tailstock end whilst you are bolting it down. If the clock starts to move you are twisting the bed, so you need some shims. When it left Myford's the bed was straight without being bolted down

It sounds as if OP is working tom take twist out of the bed, (As per Ian Bradley, and Myford's advice - "Rollie's dad's Method", so Tailstock support should not be used .

It may well be that the head bearings are slack, so that the tool is acting as an extra bearing close to the chuck

This assumes that the tool is sharp, and accurately set at centre height

The machine set up needs to be investigated and each error eliminated , to obtain a good finish.

The skill to provide a steady hand feed is one to learn.

Alternatively, the change wheels could be set up to provide the finest feed possible, about 0.004" per rev would be good target.

It will allow a better comparison of finish, since the feed per rev should then be be the same.

Howard..

The finish may improve if a very small radius is stoned on the edge of the tool. But ensure that the tool does not rub below the cutting edge...

Thanks for all the replies, I'm trying to align the lathe so am not supporting the test piece. If I do support the piece in the tailstock the finish is consistent. along the length. I am using the power feed (too much arthritis for hand feed) at the slowest speed to cut the test rings but have just realised that I think the micro-meter dials on the lathe are metric not imperial and I have been taking too deep a cut; the steel is also of an unknown type. I will try the helpful suggestions and see if I can eliminate the chatter. and poor finish at the far end test ring.

I suspect the most likely cause of the chattering/rough finish is most likely due to the chuck not gripping the workpiece firmly enough. The slightest amount of bell-mouthing (or the reverse) of the chuck jaws would have the effect of only gripping firmly at one narrow axial position along the bar rather than over, say, 30mm of its length.

Ian P

For aligning the lathe you must not use any support.

To improve the finish at the far end of the bar you could:

• lower the RPM (less resonance)
• reduce the cutting depth (less cutting forces)
• use a polished insert for aluminium (less cutting forces)
• use a thicker bar or tube (more rigid)
• use a shorter 4"bar (more rigid)
• use an aluminium bar because it is easier to turn to a good finish (less cutting forces)

Try a 4" long bar instead of 6" as per thr Myford manual.

And look for slack headstock bearings and bellmouthed chuck. Both common issues on old Myfords.

Use Rollies Dad's method ( my preference ) or glue 2 aluminium rings to the ends of the steel rod and cut the aluminium.

Edited By Y C Lui on 27/05/2022 05:28:32

I used a 6” aluminium bar for that test - I also got chatter with steel. I assumed that material wasn’t too important, and something relatively soft and easy to cut would cause less load/deflection while cutting. After all you’re not checking rigidity, just alignment.

OK, lathe is new to Morgan, who has just got the motor going.

Forgive me if I'm wrong, but I guess Morgan is inexperienced. He's started by attempting to level the bed with an ordinary spirit level and a length of unknown steel.

Various problems with this.

Lathes don't actually need to be level, but levelling is one way of tackling a twisted bed, which has various causes. An engineers level is needed to de-twist a lathe bed because ordinary levels aren't sensitive enough. A bed has to be twisted a lot before an ordinary bubble will see it. As there's no evidence yet that the bed is twisted yet, the level is confusing and might do more harm than good.

Morgan is right to question his steel, unknown scrap is a menace, not good for a delicate alignment test. I recommend buying a length of large diameter Aluminium Alloy rod where the specification specifically says it has good machinability.

Others mention problems likely to be found on a second-hand machine. Bell-mouthed chuck, spindle bearings, also loose gibs, worn saddle and/or bed, blunt tool, wonky tool post, wrong height setting etc. As these can all interfere with cutting a delicate test bar, I suggest putting the lathes alignment on the back-burner for the time being.

Start by using the lathe as is to make a number of shortish parts, less than 3 x diameter overhang out of the chuck. Learner drivers need to hone their skills as well as test the lathe. Make several parts before drawing any conclusions unless the fault is obvious. Lathe alignment and run-out don't matter much cutting close to the chuck - their ill-effects bite on long jobs, which are more difficult anyway.

Don't rush to blame the lathe when things go wrong at first because inexperienced operators make a lot of mistakes, especially if self-taught. Things like getting tool height wrong, not spotting blunt tooling, wrong RPM and feed rate, pussy-footing and bull at a gate are both bad. Many operations have a sweet spot that has to be found by practice. Take the time to build up a good picture of what the lathe does well and badly. Only start fixing when the evidence is in. With luck, the test phase won't find any significant problems. And if it ain't bust don't fix it!

Chaps often rush to strip new-to-them machines down, oil and clean everything, and take high precision measurements to prove all is well. I recommend not doing it. Stripping the lathe down prematurely risks introducing new faults due to assembly errors and maladjustments that can be hard to diagnose particularly when mixed with real problems. Lathes may not be rocket science. but they have their subtleties and accurate measuring is much harder to get right than it first appears. It's easy to get into a tangle - you can guess how I know!

When the time comes, I think Rollies Dad is better than the levelling method for aligning. Levelling seems easier to do, but I found it harder to apply in practice than Rollies. I had to read the instructions several times.

The ordinary builders level is useful for a related purpose - making sure the bench is reasonably firm and doesn't move under the weight of the lathe. If the stands legs are aren't straight and its top builders level, the whole bench is liable to twist, skew and vibrate which upsets the lathe on top. Did I mention a bendy bench can twist a lathe's bed irrespective of the owners efforts directly under the lathe's feet?

The best way to test a lathe or mill is to use it to cut metal. If a machine does that OK, it's probably fine. Diagnostic methods come into play when machines don't cut metal satisfactorily.

Dave

Dave has good points, which need addressing.

Reading through the thread, we are not told what type of cutting tool is being used - it should not be carbide (unless choosing the polished versions made especially for machining Aluminium.

Dave says:

Lathes don't actually need to be level, but levelling is one way of tackling a twisted bed, which has various causes. An engineers level is needed to de-twist a lathe bed because ordinary levels aren't sensitive enough.

This is all well and fine for lathes only secured at each end (my lathe is only securely bolted down at the headstock, with just two fixings, to keep it in position. The tailstock end is only lightly fixed in place - with a single spring-loaded bolt . This is OK for our small hobby lathes, but not OK for larger (industrial ) lathes which may require several pairs of supporting feet - which means the bed does need levelling to a precision not attainable with a cheap builders level!

As above, if you want 6” of stick-out, use at least 2” diameter stock. Checked with precision measuring kit (micrometer, not caliper), better results for parallelism can be achieved than most of us can machine to (or is needed) for most items made.

For a start, we cannot be sure that the ways are up to best industrial requirements - all the way from headstock to tailstock - particularly on mass produced lathes built to a price. Older lathes will have at least some wear since manufacture - it is inevitable and unavoidable, unless a really expensive recondioning is carried out to return the lathe to original (or better) specification.

Many thanks for all the good advice, the test piece is protruding 7 1/2 " but the collars are about 6" apart. I have increased the rpm with the potentiometer and slowed the feed to the minimum. I am using what I think is a knife tool and have applied a slight radius on the point and lubricated the test area with some thin oil and have a half thou difference between the two collars and the finish has much improved on the outer collar. I think I will leave it at that and start using the lathe (as suggested) to play with. I did receive some formal training on turning but that was over 50 years ago and I have forgotten most of what I did learn. All the tools for the Myford quick change post seem very small and delicate compared to what I remember on an industrial lathe but the lathe seems to be well constructed and I shall start to try some model building after first getting some much needed practise.

Incidentally where do you model makers buy your metal material from?

Morgan

Edited By Morgan Sweet on 27/05/2022 21:33:21

Dave is absolutely right.

If it ain't broke don't start fixing it!

With insufficient skill, knowledge and equipment, stripping a machine may change it from a reasonably precise item into something which is less accurate than when you started.

A friend bought an elderly lathe which had been stripped and "improved" by a previous owner. It would not cut parallel because a cross threaded stud was pulling the Headstock out of line. We had to bush the tapped hole so that the machine aligned correctly, and no longer cut an unwanted taper.

Older British and American machines will have been made to Imperial dimensions, and most of the threads will be Imperial, rather than Metric

Surprisingly, to me at any rate, not everyone realises the need for the tool to be mounted at centre height, and to have correct clearances.

Too low, there will be excessive front clearance.

Too high and the cutting edge will not be able to do much cutting, instead it will rub, generating heat but little or no swarf.

To prevent rubbing,a tool needs, for most purposes, to have front, and side clearances of about 5 - 10 degrees, and something similar for top rake. For soft materials, like brass, it is sometimes better to have no top rake, to minimise risk of dig ins.

For many years I have used my parting tool with no top rake, just side and front clearance, for this reason.

Excessive clearances weaken the cutting edge, and shorten tool life because of the lack of material to conduct heat away from the edge

A blunt or incorrectly ground tool will not cut well.. (A badly ground drill, if it cuts, will produce oversize holes. )

For some materials, usually tougher, or the more exotic, clearances have to be more specific to optimise the cut and finish. Ditto for speeds and feeds..

To machine mild steel, think in terms of 100 fpm (30 Mpm ), for a good finish feed should be about 0.004" /rev (0.1 mm/ rev )

No doubt others will disagree with some of this.

HSS steel will not cut hardened steel. For this carbide tools are needed

If there is slack in the bearings, or gibs or the chuck jaws are worh (Usually bell mouthed ) you will have difficulty in getting a good finish or accuracy. Similarly, tools need to be held as rigidly as possible, with minimal overhang from a solid, rigid toolpost.

As with measurements, rigidity is important. A light machine, under cutting loads, will deflect more than a heavy rigid one.. Which is why, light weight hobby machines cannot withstand operating at the speeds, feeds, and depths of cut of the very much more costly industrial ones.

Howard

Double post

Edited By Howard Lewis on 28/05/2022 10:19:58