Parting off on Myford lathes

[Alan JacksonParticipant @alanjackson47790]

I generally agree with TerryG’ s summation that a rear parting tool post works best. I have spent many hours being petrified at parting off on lightly constructed lathes with average head bearings etc. Now the rear toolpost on my chipmaster works easily without fear. I have also experimented with front and rear cutting tools and reverse rotation upside down mounted tools and believe that rear inverted cutting position works best on light lathes because the cutting tool does not seem to get to the dig in position so easily, it becomes less resonant. The cutting forces seem to reach a more stable situation and this shows up in a better finish.

Alan

Edited By Alan Jackson on 19/01/2011 17:44:19

[WALLACEParticipant @wallace]

So how about someone for the good of lathe users everywhere doing a test with two identically ground parting off tools on the same bit of steel using a rear and front mounted tool post and seeing which is best. You could steadily increase the feeds until they break

Even better, video it and stick it up on YouTube for all to enjoy –

I would offer to to the dirty deed myself – but alas – I don’t have a rear mounted tool post on my lathe . . ..

w.

[Chris GunnParticipant @chrisgunn36534]

I have read through the thread, and up to now no one has commented on the bed design of the Myford. I remember reading a long while ago in the ME that this was felt to be the reason for poor parting off in the Myford. As the bed is of rectangular construction, with no vee sections, it is difficult to get the cross slide rigid and chatter free. Parting off in a front tool post in a lathe with a vee bed surely pushes the cross slide down into the vee, and will tighten things up. When I worked at British Steel in the DRTD, we had aworkshop for small jobs, with Myfords, Boxfords and Kneller round bed lathes, everyone would pick the Knellers first to do a job, the Boxfords second, and would avoid the Myfords apart from doing a bit of polishing. I know this will upset the Myfordfans but thatis how it was.

I have a Bantam, and use front and rear toolposts, and it will part off using the feed with a sharp tool and plenty of suds, something else no one has mentioned. I was always taught to use copious amounts of suds when parting off when I did my apprentiship, ithelps keep everything cool the tool sharp and flushes away the chips.

Chris Gunn

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

(1) Elasticity of belts : Under safe conditions get someone to hold the motor pulley still and see how much you can rotate the chuck by . Make a distinction between initial taking up of the slack and the actual springiness .

 

(2) All bearings on my ML7R are usually right on at the manufacturers recommended settings .

 

(3) Belt primary drive gear head lathes are much less affected for two reasons (i) generally the more remote any elasticity is from the spindle the less effect it has .(ii) the inertia and slight damping effect of gears tend to nullify the effects of any primary drive elasticity .

 

(4) This is an experiment which has been done and which any of you could repeat with simple equipment and which always gives consistent results :

 

Set up a bar to run true and firmly in chuck and mark a radial line in high visibility paint on the outboard endface to act as an indicator . Arrange a strobe to illuminate the end face . Set up a forward facing parting tool in the normal way . Set spindle to run under power at normal running speed and sync the strobe to spindle speed . Get an assistant to do some parting cuts and observe the behaviour of the indicator line .

 

With cut not engaged the indicator line stays still .

 

When cut is engaged but cutting is stable the indicator line will precess slowly with the slight change of running speed under load but nothing much else happens .

 

When chatter is initiated the indicator line in addition to precessing will jump back and forth over a sector of several degrees . 

 

I’ll leave you all to interpret .

 

(5) A further experiment which as been done and which some of you at least could repeat and which again gives completely consistent results is :

 

Arrange a high resolution rotary position sensor on the spindle and process the output so that it can be fed into a frequency analyser .

 

When running with no cut you just get one peak at the fundamental .

 

When running with stable cutting you get the same fundamental peak with a slight shift and some non systematic noise .

 

When running with chatter you get the fundamental , more noise and most significantly another strong peak at typically 10 to 50 time the fundamentalfrequency .

 

I’ll leave you all to interpret this as well .

 

(6) If any of you have asked me questions which I may have missed please remind me .

Edited By MICHAEL WILLIAMS on 19/01/2011 20:07:05

[blowlampParticipant @blowlamp]

Michael.

 

Thanks for this very interesting reply, but in it, you seem to be concentrating on chatter, rather than the topic of parting-off.

 

When I’ve had parting-off cram-ups in the past, I don’t particularly remember chatter being part of the problem. It was more a sense of the lathe having a ‘rough’ feeling coming through the feed handle, accompanied by an unpleasant sound before the inevitable crack of the tool breaking.

 

Another comment I might make, is that if springiness in the drive system is the problem with these lathes, then I can’t see how using a rear mounted toolpost can cure it, as we will still be using that same drive system along with all it’s supposed inadequacies.

 

Did you find that using backgear eliminates the problem?

 

 

Martin.

Edited By blowlamp on 19/01/2011 20:52:17

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

The specific problem of cramup when parting off steel with HSS tools is not down to any defects in the lathe , tool setup or anything else – it is due to a bit of chemistry . Low carbon steels have a chemical afinity for Tungsten and steel and Tungsten will weldbond to each other under high pressure . HSS comes in lots of grades for different purposes but the standard Chinese imports which seem to be on sale everywhere have a high Tungsten content . So when parting off tiny bits of steel weldbond to the tool edge and after a time a little nugget forms . This spoils the cutting action completely and you end up pushing the tool harder and harder to make it cut which in turn makes the accumulation of steel on the tool worse and you end up with the well known jamup .

 

This problem can be made much better by :

 

Choosing tooling made from the correct HSS for steel cutting .

Using flood coolant .

Making sure that the basic cutting action is working well – lots of top rake and side relief .

Parting off at relatively high speed . The turning speed for parting off is the same as ordinary turning – not dead slow as is commonly employed ..

 

The Tungsten HSS tool/steel problem is the basic reason why many people who have trouble parting off have much more sucess when they change to ceramic tipped tools .

 

I mentioned a bit about the front/rear parting tool argument in my initial posting .

 

Engaging back gear for parting off modest size work probably causes as many problems as it solves . In particular using too low a speed means that your relative feed is all wrong and you are generating high tool loads for no gain .

Edited By MICHAEL WILLIAMS on 19/01/2011 23:26:17

[Nicholas FarrParticipant @nicholasfarr14254]

Hi,

Tony, I haven’t said that belts are not elastic quite to opposite if you read my first post. My “stretchy”terminology was obviously not understood, what I ment was akin to a soft rubber band where it will be obvious that it is significantly stretching under load and then relaxing when the amount of load decreases. While this does happen, it is not significant enough for the operator to notice, being so small when the belts are driving. The driving side of the belt is always in tension during rotation, while the slack side is constantly taking up the differance in the whole system by the phenomenom known as certrifugal force.

 

Michael, you can’t compare a static situation of holding the motor pulley still while seeing how much your chuck will turn, with a rotating system for the reasons I have metioned above. You can’t introduce the forces that keep the drive balanced.

 

These are my opinions based on being hands on involeved with belt drives for all my working life.

 

Wallace, I was thinking filming the events might shine some light on the situation, but I think you would need high speed filming in conjunction with known refferance points, and maybe data from strategically placed strain gauges.

 

Regards Nick.

Edited By Nicholas Farr on 19/01/2011 23:26:53

Edited By Nicholas Farr on 19/01/2011 23:27:34

Edited By Nicholas Farr on 19/01/2011 23:30:36

[Ian S CParticipant @iansc]

Although the V belt will fractionally stretch, I feel that most of the apparent stretch is in the fact that as the load goes on, the belt pulls down in the V (the belt normally should not bottom in the V). I use link type belts, and have no problem. Ian S C

[KWILParticipant @kwil]

Have we now come to the conclusion that it is nothing to do with it being a Myford or not, merely down to technology, materials and technique?

[John Stevenson 1Participant @johnstevenson1]

Getting back to a Myford.

Posted by Michael Williams “(1) Elasticity of belts : Under safe conditions get someone to hold the motor pulley still and see how much you can rotate the chuck by . Make a distinction between initial taking up of the slack and the actual springiness .”

 

 

Better still chuck a decent sized bar up 3/4″ or 1″ about 10″ long protruding from the chuck , put a dial gauge on the bar about 2″ out, roughly where you work the most, with the base on the cross slide.

 

Now grab the end of the bar and wobble it about and watch the gauge.

 

I’ll leave you to draw your own conclusions and none of these relate to belts, gears etc

 

 

John S.

Edited By John Stevenson on 20/01/2011 09:49:39

[The Merry MillerParticipant @themerrymiller]

Can I take it as read that everybody ensures that the parting-off tool is dead at right angles to the headstock spindle axis and the saddle is locked before the tool starts cutting.

I may well have missed these points in earlier posts, please forgive me if I have!

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

(1) KWIL : Refererences to Myford are only because its construction is familiar . I would expect most smaller lathes of similar construction to have the same problems .

 

(2) IANSC : Quite right – belt settling into the pulleys under load adds to the elasticity of the system .

 

A man somewher converted his Myford to all PolyV drive and reported much reduced parting off and chatter problems . This is reasonable since PolyV belts fit the grooves in the pulleys much better than ordinary V belts and since PolyV belts run with relatively high tension the elasticity in the system is much reduced .

 

PolyV belts are one trade name of a type of belt which is flat and thin and which has multiple Vsections on the drive side – looks like a row of mini V belts all stuck together side by side . These match multi V grooves in the pulleys .

 

(3) I would really like someone other than me to decode those experimental results and explain what they are saying .

Edited By MICHAEL WILLIAMS on 21/01/2011 10:30:56

[Tony JeffreeParticipant @tonyjeffree56510]

Posted by MICHAEL WILLIAMS on 21/01/2011 10:17:44:

(3) I would really like someone other than me to decode those experimental results and explain what they are saying .

OK…I’ll bite.

The experiments you describe are basically measuring (changes in) rotational speed of the spindle/chuck/workpiece. In the first one (strobe), you set the strobe so that it fires exactly once per revolution; the indicator line appears to be stationary because he strobe light catches it in exactly the same position on each rev.

Applying cutting force to the work will slightly slow the rotational speed of the spindle, so the indicator line will appear to rotate slowly clockwise when viewed from the tailstock end, because the spindle is taking longer for each full rev than the interval between strobe pulses, so each pulse “catches” the indicator line slightly too soon. (Hence, clockwise rotation shows speed rediction, and anticlockwise would show speed increase, which you could easily demonstrate with a variable speed drive.)

In the chatter case, the average speed is still reduced, but because the motor/drive train/spindle assembly is now resonating (i.e., the instantaneous rotational speed is varying above and below the average rotational speed) because of the elasticity of the drive train, the indicator appears to jump back and forth either side of the “average” position.

In the second experiment, the explanation is basically the same, and what it is showing is (a) the average speed of rotation (the first peak) and (b) the frequency of the resonance in the drive train (the second peak).

John S’s experiment was pointing out that the workpiece/chuck/spindle assembly is itself flexible, and can therefore also be made to vibrate in other modes under the right conditions, and that will also induce chatter. However, if you eliminate that factor from the equation (large diameter bar, cut close to the chuck, massive/rigid tool/toolpost), then you can still create chatter via the kind of rotary oscillation that your two experiments illustrate.

The fix for the rotary oscillation problem is a stiffer drive system and/or more “flywheel” mass at the spindle end of the drive train and/or more damping.

Or you simply follow the old maxims…”reduce speed, increase feed” etc…to shift the system away from its resonant frequencies.

Incidentally, conducting your second experiment with different sizes of chuck but nothing else in the system changed would show a change in resonant frequency; the smaller the chuck, the higher the frequency of that second resonance peak. Similarly, using a stiffer belt would increase the resonant frequency.

Regards,

Tony

Edited By Tony Jeffree on 21/01/2011 11:07:24

[blowlampParticipant @blowlamp]

Right, second bite…

So what we’ve established so far is that chatter is always present, but with care in the design of the machine, tooling and by using appropriate speeds etc, we can tune it out of harms way and so produce good work.

 

This is my take on what happens and is what I believe causes parting-off problems with the commonly available HSS blades when used in the front toolpost:

 

1/ The cut commences and is initially fine.

My View Is That:-

Depite the tool being quite wide and so removing more material than in normal turning, the basic Machine/Tool/Speed setup is adequate for the job.

 

2/ Once a certain depth is reached, sensations can be felt through the lathe and the sound of the cut can change. At this point I would say a lot of people instinctively withdraw the tool, clean the groove and carry on.

My View Is That:-

This is the point that the system is beginning to become unstable.

 

3/ A short while after the cut is recommenced, the cram-up happens – but why?

My View Is That:-

The parting-off blade is now working in a confined space with very, very little clearance.

As various sized chips are generatedby the cutting action, most of the larger ones are ejected by other swarf as it is being produced. However, small, gritty size pieces can remain and some of these will become lodged between the tool and the workpiece and might even ‘pick-up’ or ‘gaul’ the job.

From there, it can be seen that this unstable system is working in a situation where a log-jam can now occur as other swarf is produced.

When the log-jam does happen, the system experiences large forces and from these forces deflections can be seen about the tool and it’s mounting, as well as the visible signs of the work moving too.

 

So to work around the problem:-

Some folks will widen the groove, before carrying on with the first cut as it gives more room for the swarf to clear.

Some will part-off from the rear, which allows gravity to help keep the groove clear of swarf.

Others will use a specially designed ‘tipped’ tool which curls the chip to make it narrower and also has plenty of clearance at the sides.

Etc.

 

Martin.

Edited By blowlamp on 21/01/2011 12:30:41

Edited By blowlamp on 21/01/2011 12:31:35

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Tony Jeffree – many thanks for your explanations which were clear and certainly correct .

 

There are resonances and more simple instabilities to be found in any machine tool design and there are often cases in which cutting loads can act in different ways according to the specific tooling and workholding set up in use at any one time.

 

A successful machine tool design usually has massive rigidity but perhaps as a result of these discussions it has become more clear that rigidity is not simply saying tons of metal but it is something more subtle requiring clear understanding of how forces are transmitted through the whole of the fixed and moving components .

 

Some sophisticated high accuracy CNC machines have active elecronic damping where any instabilty is detected and feedback is sent to the main and axis motors to compensate .

 

Blowlamp – part of the problem of log jamming and resultant tool breaking is human response . Most people when the tool action is blunted for whatever reason turn the feed handle a bit harder to compensate and this initiates an immediate downhill spiral to disaster . When you feel or suspect loss of free cutting action retract the tool immediately and clean or sharpen tool .

 

Some turners when doing more hazardous parting off or deep grooving use a controlled feed /retract cycle as normal  – ie feed in to a certain depth , retract tool to clear , feed in another increment , retract and repeat as many time as needed .

 

Another minor detail is that most people make the cutting end of the parting tool the wrong shape . The conventional simple flat ended shape with sharp square corners

is not the best shape for reliable parting off . What happens is that the extreme corners can get broken off or blunted very early on in the cut and you are now trying to drag an effectively blunt tool through the workpiece . A far better shape is one which is flat ended but with the outer corners rounded (and given relief under) . You may get a small extra amount of pip on the cut off pieces but your tool breakage will drop considerabley .

Edited By MICHAEL WILLIAMS on 21/01/2011 13:21:15

[blowlampParticipant @blowlamp]

If anyone needs further evidence of the effect that chip crowding has, you only have to drill deep holes in certain materials such as aluminium to see how easily it can jam the bit.

 

Another example is if you try to deepen an already threaded hole at core size. In this instance the swarf is forced back down the hole as it rotates with the bit and is grabbed by the thread section of the hole.

 

Both are easy to verify for oneself.

 

Martin.

 

Edited By blowlamp on 21/01/2011 14:00:54

[KWILParticipant @kwil]

Martin and Micheal’s conclusions are the same, use the correctly designed tip tool which curls the chip narrower and incidently have radiused corners of between 0.10 & 0.30mm (Sandvik), which is what I use.

[TerrydParticipant @terryd72465]

Hi Blowlamp,

You can’t really make an analogy between a twist drill and a parting tool. Any jamming with a drill is due to completely different processes. The swarf from a drill will curl freely in the small space of the flute in a helical form, especially when cutting aluminium. This will tend to block the free passage of the swarf as the tool gets deeper. This effect is compounded by the fact that the flute is deepest at the cutting point and gets shallower towards the shank, therefore the room for the chip to exit is is getting smaller, hence crowding. The answer is peck or intermittent drilling which breaks the swarf into shorter lengths and prevents the initial curling of the swarf.

A properly sharpened and set parting tool produces a spiral not a helix due to a different cutting angle compared to a drill point and it’s flute, the chip is also the same width as the cutting tool. If you see chip crowding in a parting groove it is very easy to remove with a scriber. It is unlikely that this will cause a 15 0r 20mm wide hss tool to break. if properly set. If you inspect a parting groove following a tool breakage you will see the relatively deep gouge where the tool has dug in.

Best Regards

Terry

[blowlampParticipant @blowlamp]

Hi Terry.

Just to clarify the position. My last post was only intended to reinforce my point about the effect chip crowding can have with regard to it’s catastrophic consequences, rather than be a direct comparision with parting-off in the lathe.

I don’t quite follow yourpoint about seeing and removing chip crowding from the groove, because my experience (in the dim and distant past of course) is it’s pretty well over by then

WOW!! Is that a 15-20mm wide parting-off tool you mention towards the end

Martin.

[Alex gibsonParticipant @alexgibson50133]

Hi All,

I don’t have any experience of Myfords, I have a Sieg C4 and have experienced all the scary stuff when parting off and have often wondered if the problems are more to do with the surface speed of the work passing the tool tip. If for example I try to part off a workpiece and set everything up to cut sweetly at first. As the tool is fed into the workpiece the rpm and feed rate are constant however the actual speed of the metal passing the tool tip decreases as the diameter gets smaller and sooner or later I get the “submarining effect” that TerryD described before. So at some point in the cut should i increase the rpm or slow down the feed to maintain the same surface speed?. My pea brain has been contemplating this for some time now, but I can’t get to a conclusion. So parting off in my shop (garage) is usually done with a hacksaw, or if it”s bigger than 10mm the bandsaw.

alex

ps sorry i if I’m being too basic for this thread.

[John OlsenParticipant @johnolsen79199]

I’m not too convinced that the poly V belts make much difference to parting in the Myford, I have done this conversion to my ML7. It is a worthwhile thing to do for many reasons, but I don’t think it has improved the parting off performance. Which incidently lately has been better than it was before I realised that the spindle bearings were due for an adjustment. I think it is still not as good as my Unimat 3 acheives, but then, the Unimat does not attempt quite such ambitious sizes, and is not nearly so old too. (The Unimat is driven by toothed belts. ) Incidently the Poly v conversion was described in ME for the Super 7 and Hemingway sell the bits for it, although mine didn’t come from there. It can with a bit of ingenuity can be applied to the ML7. I managed to give myself four belt speeds instead of three, with the top and bottom being much the same as the original, so the ratios are close. Probably superflous because I also have the Newton-Tesla drive on it.

Somebody said something to the effect that chatter was always present. I would say rather that the potential is always there, in that you have a system, or rather a number of systems, with the potential to oscillate. Drawing on my electronic background, such a system will not oscillate unless the energy input is sufficient to overcome the losses. There is usually some sort of loss, in an electronic system it will be resistance, and in our system it will be damping due to friction. If the damping is sufficient then oscillations will not be sustained. You can sometimes see this with an interrupted cut, where there may be a little sign of chatter just where the cut starts, but the rest of the cut is fine.

Alex, with a small machine like the Sieg, I would suggest trying the narrowest parting blade you can manage. I have used one just over a mm wide on the Unimat with good success. This is ground from a piece of 1/4 inch HSS, so takes a while to make, and is long enough to part off up to about 5/8 diameter. The usual commercial blades are just too wide for these small machines. You can apparently make a holder and use a piece of hacksaw blade to make a tiny parting tool, but I haven’t tried that myself.

regards

John

[TerrydParticipant @terryd72465]

Hi Martin,

 

If you didn’t intend to compare the two why use the comparison? It seems disingenuous to me. I’m sorry but so called ‘chip crowding’ takes place over a relatively long period of the process (several seconds) I’m surprised that you are not aware of this so called phenomenon happening. By the way, if chips are jamming so firmly, how come they can simply ‘fall out’ when a rear toolpost is used, by the simple effect of gravity?

 

As to your last point – with all due respects, 15 to 20mm depth of HSS parting tool, which is the normal size as most engineers are aware, I assumed that someone as perceptive as yourself would understand that without it having to be explained.

 

Best regards

 

Terry 

Edited By Terryd on 21/01/2011 22:49:13

[TerrydParticipant @terryd72465]

Hi Martin,

 

By the way, you can try the chip removal with a scriber or similar tool after the parting failure, the chips should still be there if they are so badly jammed. The only chip which is difficult to remove is the one produced when the tool has finally jammed.

 

Kind Regards

 

Terry

Edited By Terryd on 21/01/2011 23:21:36

[Nicholas FarrParticipant @nicholasfarr14254]

Hi Tony and Michael,

             while performing your strobing experiment, I’ll assume that you were strobing the motor pulleyat the same time and found that there was no oscillation in your timing mark there then, and/or suitable calculated for the two differant speed ratios.

 

Regards Nick.

Edited By Nicholas Farr on 21/01/2011 23:28:33

[blowlampParticipant @blowlamp]

Terry.
To answer your last posts.

 

“If you didn’t intend to compare the two why use the comparison?”
I didn’t compare drilling with parting-off. I simply used drilling as an easily verifiable example of the effect of chip crowding i.e., the drill bit stops turning when jammed with swarf.

“It seems disingenuous to me.”
Read the post again and you should see that it isn’t.

“I’m sorry but so called ‘chip crowding’ takes place over a relatively long period of the process (several seconds) “
Really, so why does it catch out so many people and result in broken tooling and damaged work?

“I’m surprised that you are not aware of this so called phenomenon happening.”
I am, I was the one that mentioned it first.

 

“By the way, if chips are jamming so firmly, how come they can simply ‘fall out’ when a rear toolpost is used, by the simple effect of gravity?”
The majority of them fall out of the way because there is no parting-off blade underneath them to block their exit.
When compared to a front mounted parting-off tool setup, it’s obvious that swarf will rest upon the upper surface of the blade by the effect of gravity and by that same effect, along with vibration, the smaller particles can tend to collect in any gaps that are present, thus magnifying the chances of a log-jam style cram-up.

 

“As to your last point – with all due respects, 15 to 20mm depth of HSS parting tool, which is the normal size as most engineers are aware, I assumed that someone as perceptive as yourself would understand that without it having to be explained.”
Well you wrote 15 to 20 mm wide and knowing what a stickler you are in your writing, I took you at your word. You made a straightforward mistake, but instead of correcting it, you’re trying to put me on the spot for asking if what you wrote, was what you really mean to say – now that’s disingenuous.

Martin.

Edited By blowlamp on 22/01/2011 00:58:54

[Author]

Posts