Lathe facing convex or concave

[BazyleParticipant @bazyle]

Or should it face dead flat

[BazyleParticipant @bazyle]

The 'what did you do today' thread is getting a bit waylaid with discussion of whether a lathe should face flat or convex. So I thought discussion could continue here as it may be of interest to others doing a search in the future. For the intitial part see the the 'What did you do today' thread around 24 Feb 2014.

My Hobbymat saddle design and adjustment is so awfull it can be made to face either way. Needs keepng an eye on when it matters.

Otherwise I suggest power feeding will tend to wear a saddle such that it rotates anticlockwise viewed from above so over time it will face off more convex.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Bazyle on 24/02/2014 20:23:50:

The 'what did you do today' thread is getting a bit waylaid with discussion of whether a lathe should face flat or convex. So I thought discussion could continue here as it may be of interest to others doing a search in the future.

.

An excellent idea, Bazyle.

To my mind, the answer is very simple: If the manufacturer could achieve everlasting perfection, then the lathe should be made to face flat. In the real world, however, there are are tolerances for everything and Dr. Georg Schlesinger was right. …

For it to be useful; a lathe must face somewhere between "slightly concave" and "dead flat". How close to flat therefore depends upon how accurately made, and how resistant to wear the lathe might be.

MichaelG.

Edited By Michael Gilligan on 24/02/2014 20:58:43

[Roderick JenkinsParticipant @roderickjenkins93242]

Tubal Cain wrote extensively on lathe precision and setting up in ME Vol 177 (Nov-Dec 1996).

Perhaps these articles could be repeated or we could have access via the website?

Rod

[Neil WyattModerator @neilwyatt]

Can anyone guess who wrote:

"Accuracy is a feature much to be desired in any machine tool, and the reputation of a lathe may stand or fall by its virtues in this respect. In recent years tstandards of accuracy for machine tools, including the well-known Schlesinger limits, have been set up to define the essential factors of concentricity, alignment etc. which are consdierd desirable in machines. Generally speaking, however these can only be observed in high class industrial machine tools; the types of lathes which have to be built at a competitive price are often sold without any guarantee as to accuracy … In the past, many small lathes have been manufactured in which accuracy was very dubious, but in view of their low cost, their imperfections could be tolerated, and they served the purpose for which they were intended quite well. It does not follow that accurate work is only possible if the machine tool itself is above reproach … the requirements of the amateur, to whom time is of less importance, may be adequatel;y served by a simpler and less expensive lathe."

And

"Lathe users, other than those with practical experience of machine tool fitting, are strongly advised not to attempt any modifications which might permanently affect the accuracy of a lathe. It is safer to accept the inaccuracies of the machine and learn to live with them. Even the most expensive lathe cannot be expected to remain in perfect adjustment indefinitely, and all machines have their limitations."

Neil

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

And all milling machines and surface grinders should have banana shaped table slideways for the same reason ?

[NobbyParticipant @nobby]

Check my 40 year super 7 mark 1 turns measure's 0.0003 " over "

Edited By Nobby on 24/02/2014 23:11:52

Edited By Nobby on 24/02/2014 23:12:27

Edited By Nobby on 24/02/2014 23:12:47

[NobbyParticipant @nobby]

Should have said concave over 3"
Nobby

[John Stevenson 1Participant @johnstevenson1]

Posted by Neil Wyatt on 24/02/2014 22:03:21:

Can anyone guess who wrote:

"Accuracy is a feature much to be desired in any machine tool, and the reputation of a lathe may stand or fall by its virtues in this respect. In recent years tstandards of accuracy for machine tools, including the well-known Schlesinger limits, have been set up to define the essential factors of concentricity, alignment etc. which are consdierd desirable in machines. Generally speaking, however these can only be observed in high class industrial machine tools; the types of lathes which have to be built at a competitive price are often sold without any guarantee as to accuracy … In the past, many small lathes have been manufactured in which accuracy was very dubious, but in view of their low cost, their imperfections could be tolerated, and they served the purpose for which they were intended quite well. It does not follow that accurate work is only possible if the machine tool itself is above reproach … the requirements of the amateur, to whom time is of less importance, may be adequatel;y served by a simpler and less expensive lathe."

And

"Lathe users, other than those with practical experience of machine tool fitting, are strongly advised not to attempt any modifications which might permanently affect the accuracy of a lathe. It is safer to accept the inaccuracies of the machine and learn to live with them. Even the most expensive lathe cannot be expected to remain in perfect adjustment indefinitely, and all machines have their limitations."

Neil

.

.

You did, just now.

[Roderick JenkinsParticipant @roderickjenkins93242]

Posted by Neil Wyatt on 24/02/2014 22:03:21:

Can anyone guess who wrote:

"Accuracy is a feature much to be desired in any machine tool, and the reputation of a lathe may stand or fall by its virtues in this respect. In recent years tstandards of accuracy for machine tools, including the well-known Schlesinger limits, have been set up to define the essential factors of concentricity, alignment etc. which are consdierd desirable in machines. Generally speaking, however these can only be observed in high class industrial machine tools; the types of lathes which have to be built at a competitive price are often sold without any guarantee as to accuracy … In the past, many small lathes have been manufactured in which accuracy was very dubious, but in view of their low cost, their imperfections could be tolerated, and they served the purpose for which they were intended quite well. It does not follow that accurate work is only possible if the machine tool itself is above reproach … the requirements of the amateur, to whom time is of less importance, may be adequatel;y served by a simpler and less expensive lathe."

And

"Lathe users, other than those with practical experience of machine tool fitting, are strongly advised not to attempt any modifications which might permanently affect the accuracy of a lathe. It is safer to accept the inaccuracies of the machine and learn to live with them. Even the most expensive lathe cannot be expected to remain in perfect adjustment indefinitely, and all machines have their limitations."

Neil

George Schlesinger?

[John Stevenson 1Participant @johnstevenson1]

Ironic though really when you think about it is that the old guy who taught me machining and had been in the job all his life, First job he ever did was rudder bolts on the arc, last job was brass turnbuckles for his coffin.

 

If you had asked him who Schlesinger was he would not have known.

 

Now if you had asked him how to tweak a hole 0.0001" either way with a reamer he'd have known.

 

I wonder if "Borderline Schlesinger" is another name for OCD ?

Edited By John Stevenson on 24/02/2014 23:31:53

[BazyleParticipant @bazyle]

Just realised the last word on my first post should have been 'concave'.
We often get threads where someone has a grumble about Chinese lathes shortcomings but we don't normally get anyone mention this aspect. Is it just they don't know about the possbility of error or are imports particularly good in this respect?

[Roderick JenkinsParticipant @roderickjenkins93242]

Posted by John Stevenson on 24/02/2014 23:31

I wonder if "Borderline Schlesinger" is another name for OCD ?

Probably. I've no idea whether my lathe turns concave or convex but I have set it up so it turns parallel because that is important to the sort of model making I do. I sometimes wonder whether the people who obsess about these things actually make anything.

Rod

[I.M. OUTAHEREParticipant @i-m-outahere]

My C2 clone is highly adjustable in the headstock department !
The" V" on the bed doesn't even come close to touching the "V "groove in the headstock and I can easily slide an .008 inch feeler gauge between the surfaces !

So the only thing stopping it moving is the tension of the mounting bolts .
Looks like I will get to play in my box of shim stock to fix this and I think it will be tedious at best or I could machine the headstock down followed by an adjustment to the tailstock – more risk and still tedious or maybe roll it on its side drill and ream for some taper pins (after setting the head to turn parallel ) – uncertain if it would work or I could just not worry about it – I think my slightly larger lathe is the same and it has never moved on me even with the worst parting tool dig in I have ever had (prior to rear toolpost and insert type parting tool) .

I was starting to think that the only purpose that "V" groove had was to stop one putting the headstock on backwards !

I was always taught to set up a lathe to turn parallel as it will mean that anything you bore will hopefully have a bore with parallel sides also which in turn covers two of the three most common lathe operations – boring , turning and facing , ideally with the lathe set to turn and bore parallel then it should face flat or very slightly convex.
If it doesn't then the cross slide is out of kilter in relationship to the travel of the saddle along the bed .

Ian

[John McNamaraParticipant @johnmcnamara74883]

A lot of the work we do on a lathe is not between centres. Allowing the deflection demon to spoil our work even a say 20mm shaft will deflect away from the cutting forces. if it does the work will be a larger diameter as the cutting point moves towards the tailstock. Very annoying if you are fitting a shaft to an existing hole by feel. It will appear to be tight but when you finally get it to enter the hole it will be loose, Not good.

On the other hand if you are turning a base of some sort that has to sit on a flat surface you want it to be touching the surface all around its periphery not possible if it is slightly convex.

Both the above cases suggest that making the lathe turn very slightly concave is better then allowing it to turn convex.

Edward F connely In "Machine tool reconditioning" 530 pages

Recommends (For a tool room quality lathe) the end of a 12 inch test bar points .0000 to .0003" towards the operator

Boring is a different case you want it to be dead on however that is almost impossible, almost because. a few degrees change in the temperature can move the whole lathe frame out of alignment. I guess that is why many books on the subject suggest we err slightly away from perfect in the direction that is most likely to help us produce good work.

 

Regards
John

Edited By John McNamara on 25/02/2014 08:25:20

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Just a few thoughts :

(1) When facing large diameter work on an ordinary lathe the natural cutting action often causes the surface to be non flat anyway .

(2) In real world engineering a component to be faced would have a flatness tolerance . This could be non specific ie just reasonably flat or very demanding with multi parameter dimensional and geometric tolerances .

As the flatness requirement gets more and more demanding then more and more sophisticated machines and machining methods are needed .

(3) I’ve heard the convex surfaces bedding down argument many times but just don’t see it myself . The better you want the bedding down to be the more flat the surfaces have to be .

(4) Moderb CNC machines don’t have any concession to funny alignments of spindles and slides . As far as it can possibly be achieved such machines are in perfect alignment on all axes so that all milled or faced surfaces come off dead flat .

Regards ,

Michael Williams .

[Michael GilliganParticipant @michaelgilligan61133]

Posted by John Stevenson on 24/02/2014 23:31:25:

Ironic though really when you think about it is that the old guy who taught me machining and had been in the job all his life, First job he ever did was rudder bolts on the arc, last job was brass turnbuckles for his coffin.

If you had asked him who Schlesinger was he would not have known.

Now if you had asked him how to tweak a hole 0.0001" either way with a reamer he'd have known.

I wonder if "Borderline Schlesinger" is another name for OCD ?

.

John,

Pithy, and very apt … My father was much like your tutor.

I have the greatest admiration for those who can do "fitting", and for the watchmakers that make [by hand turning] parts smaller than the chips of swarf we usually produce … But: the point is, surely, that Schlesinger was prescribing limits for Machine Tools.

MichaelG.

Edited By Michael Gilligan on 25/02/2014 09:16:11

[John Stevenson 1Participant @johnstevenson1]

Posted by MICHAEL WILLIAMS on 25/02/2014 08:59:07:
Just a few thoughts :

(4) Moderb CNC machines don't have any concession to funny alignments of spindles and slides . As far as it can possibly be achieved such machines are in perfect alignment on all axes so that all milled or faced surfaces come off dead flat .

Regards ,

Michael Williams .

True they don't have any fancy alignment concessions but they have something better than that – G Code

You run your sample part with the feeds and speeds and most importantly, depth of cut, you are going to use and if you have a slight taper which is not unusual then you tweak the code to turn a taper that corresponds.

This way it's special to this job, this material and as i mentioned before, depth of cut and you don't have to alter the machine. In fact you can even compensate for a worn machine.

And whilst we are on the subject of armchair machining a-gain [ best Forrest Gump voice ]

The British Standard for surface finish is the Ra, however this isn't used in normal day to day language like thou, or knob on.

The normal day to day standard is the Badgers Arse, as in " rough as a Badgers arse " but how does this compare to the actual British Standard ?

Can anyone give a definitive answer what Ra is a badgers arse ?

[Involute CurveParticipant @involutecurve]

Posted by MICHAEL WILLIAMS on 25/02/2014 08:59:07:
Just a few thoughts :
snip<

(4) Moderb CNC machines don't have any concession to funny alignments of spindles and slides . As far as it can possibly be achieved such machines are in perfect alignment on all axes so that all milled or faced surfaces come off dead flat .

Regards ,

Michael Williams .

This is true however even high end machines have adjustments, for instance to ball screw mapping and bed slope etc built into the software and setup at the factory, very small adjustments but still there, keep in mind, new machine tools are made by old ones……..

[SaxalbyParticipant @saxalby]

Posted by John Stevenson on 25/02/2014 09:16:17:

The normal day to day standard is the Badgers Arse, as in " rough as a Badgers arse " but how does this compare to the actual British Standard ?

Can anyone give a definitive answer what Ra is a badgers arse ?

I am trying to imagine a couple of BSI engineers holding down a badger ( vicious little buggers) while running a surface gauge over its rear end.

Barry

[blowlampParticipant @blowlamp]

Posted by John Stevenson on 25/02/2014 09:16:17:

Can anyone give a definitive answer what Ra is a badgers arse ?

I don't know at the moment, but I am looking into it.

Martin.

[Ian S CParticipant @iansc]

Neil, I wonder if it was someone like Whitworth or someone of that time that made your first quote.

With a milling machine the table is going to bend like a beam, but then we plonk a wacking great rotary table, or some such bit of gear in it, and upset it some more, specially with some of our light weight bits of equipment.

Ian S C

[Anonymous]

In Euclidian geometry a plane can be defined by three points (provided they are not colinear), no straight lines, or calculus, needed. Since a plane is a 2D construct it must be flat, by definition, as there is no third dimension. Of course in the real world, which is inherently at least 3D, the concept of flatness has a tolerance.

As for badgers I think that BSI gave up trying to equate Ra to BA as there are too many variables, like culled or not, road kill or not, and if so what size vehicle. I believe that Rz, a peak to peak measurement, is now the preferred measure of surface roughness. Ra is an average value and can hide a multitude of sins, depending upon the crest factor of the surface.

Andrew

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

(1) I’m familiar with in cycle compensation etc for machining errors but as JS says these are job specific .

This is very different from setting a manual lathe up crooked in the hope that now and again something will turn out right by chance .

The manual lathe equivalent of compensation is putting jacks under it and twisting it to a new shape for each job ?

Actually the skilled machinists I grew up with ‘ knew their machines ‘ and apart from always knowing exactly what feeds , speeds and tooling to use they would often apply their own version of in cycle compensation – they would lean on the slides to give slight biases in favourable directions .

(2) It is not possible to quantify roughness fully just using simple indicator values .

Ok for bacon slicers maybe but any more sophisticated work has to have many things into account .

One feature of roughness that is not immediately obvious is that it often has directionality – usually coming from the machining process . This is why it is important to specify the lay or machining pattern on some components .

Fascinating things to study are the linked subjects of flatness , roughness , contact theory , bearing theory and lubrication theory – especially when looking at really flat and really smooth surfaces for special applications .

Regards ,

Michael Williams .

[Michael GilliganParticipant @michaelgilligan61133]

"Scratch and Dig" seems a very pertinent measure for the Badger.

MichaelG.

Edited By Michael Gilligan on 25/02/2014 12:03:46

[Author]

Posts