View Full Version : Turning cubes on a faceplate

Al Messer
12-17-2003, 10:21 PM
Since it is a "given" that work produced on the faceplate will be slightly concave, just does one go about producing Steel cubes with flat faces and square corners? Don't tell me to use the mill 'cause I ain't got one and am too broke to go and buy one or I would.

12-17-2003, 10:48 PM
It depends on how flat the cube needs to be. The specification for Sheldon lathes is to face hollow or concave only on 10 inch diameter 0 to 0.0008".
That is not enough to worry about in most cases.

12-17-2003, 10:57 PM
Al: I've commented on this (concavity) a while back based on what I had read in an old (1937) book by Smith from MIT. Evan also commented on this. I have to admit that I've never checked for it on parts I've made in the past.

I'm starting to believe that it may be an old wives tale. Several machines I've checked recently showed only a few tenths of deviation across a faceplate. Also, the machine reconditioning book which many bought from Ken G. shows the method for measuring the perpendicularity to assure that it is met.

Anyone care to comment on this phenomena (or lack of it)?


12-17-2003, 11:05 PM
As I pointed out, the dimension is very small to nothing.
No dimension on a machine tool is going to be perfect zero zero. The more care taken in manufacturing and fitting, the closer the ideal may be approached, but the idea is to be out in the direction which will cause the least problems.

12-17-2003, 11:09 PM
Concavity on a facing operation I have read something about this and from what I remember it has more to do with the facing not being a constant speed operation I.E.the surface speed of the cutter not being constant over the face of the cut.

Several lathes have been made to over come this and quite a few were really conplex,but I think with the use of modern methods like VFD'S it should be simple.

12-17-2003, 11:20 PM
I've heard the concavity discussion for years. Way I remember (from some old book) was that IF you were unable to get it flat, be sure things were slightly concave, otherwise the surface would (could?) rock. There are only three posible conditions- flat, concave, convex. Flat is impossible (in practice), so make it concave if possible

Just one mans opinion again.

12-17-2003, 11:24 PM
As Steve says, concave is way better for most fits than convex. Also as the tool approaches center pressures that tend to spring the tool build. It should be aligned to turn ever so *slightly* concave. The result will be very close to flat.

12-17-2003, 11:37 PM
The tolerance for facing is 0-.0005" for toolroom lathe, and 0.001" for 12" to 36" engine lathe, concave on 12" diameter,

C. Tate
12-18-2003, 12:03 AM
Put a fly cutter in the chuck and use cross slide as a mill table. May be a little slower but it will be same as a horizontal mill that way.

12-18-2003, 04:11 AM
You grind or scrape them square. Just like all toolmakers have done in the past when they needed cubes.

[This message has been edited by Thrud (edited 12-18-2003).]

12-18-2003, 07:15 AM
Excuse me guys, but this has to be one of silliest concepts I've read about in a long while. Lathes made to face parts so the cut area is a concave surface???? Does the term "old wives tale" ring a bell?

If you're worried about a part rocking when supported on the face, turn a recess near the center.

12-18-2003, 09:25 AM
Since it is specifically called out in the specifications of the manufacturer of a quality made lathe, it can hardly be called an "old wives tale".
It is done, and for a very specific reason. Making manufacturing tolerances work in your favor is an important item to be considered in the design of any machine.
As the relative speed of the part is decreasing as center is approached, the cutting forces will decrease, not increase. This will reduce the tendency for the tool to dig in. There would be more of a tendency for the part to be convex than concave.
It may be a bit of a stretch, but, when you think about it, the most precision made, hand scraped, surfaces are not flat, but are a series of concavities which remove the errors.

12-18-2003, 12:28 PM
"Since it is specifically called out in the specifications of the manufacturer of a quality made lathe, it can hardly be called an "old wives tale"."

What lathe is that?

This concept could get kind of tricky on a gang tooled lathe. Use an upside down tool on the back of the slide and you get a convex surface, use a rightside up tool on the front and you get a concave surface?

Until further proof I'm not buying the whole idea. And writings in a 1937 book certainly don't have application in today's higher precision machine tools.

I would call someone at Hardinge and ask except I'd feel like a complete doofus even posing the concept to them. Incidently, you can get access to machine designers at Hardinge. I have one of their older automatic screw machines (1971 vintage), when I had a problem with the control relays I called service and was connected to one of the engineers who had worked on the machine's logic design.

12-18-2003, 01:17 PM

As center is approached the work is approaching the tool cutting edge at an increasingly oblique angle (read: wrong angle) which increases pressures.


Chris Heapy seem to think that lathes face slightly concave. Search on the word concave on this site.


I believe it is mentioned in the book "How to Run a Lathe" as well. I'll have to look.

12-18-2003, 01:49 PM
"Chris Heapy seem to think that lathes face slightly concave. Search on the word concave on this site.


I believe it is mentioned in the book "How to Run a Lathe" as well. I'll have to look. "


I checked the site you mention, didn't see a search option. "How to Run a Lathe", isn't that the oldish book published by South Bend or was it Atlas?

Several posters on this subject (here and elsewhere) say their lathes exhibit this concavity facing phenomena. I don't argue with what they're seeing. Some seem to have extended this observation to conclude their lathes were designed to do this. I suggest what they are seeing is more due to lack of rigidity in their machines causing the cutter to dig in deeper as the tool's angular approach to the cut changes with decreasing diameter and decreasd cutting speed.

If there are individuals here that truly believe lathes are purposely designed to do this concave business then I ask them where does it stop? Is this only designed into small home shop lathes, or does it extend also to very heavy rigid lathes and the super precision CNC lathes. How about the lathes we use to turn high precision lens blanks, I don't recall any programming precautions we have to take to compensate for inherent concavity-ness of the machines.

Again, I think this is just some old wive's nonsense.

12-18-2003, 02:04 PM

Use Edit, Find on this page from the menu bar.

It is like the specs for precision ball bearings. The outer shell OD is allowed to be .0000 oversize and .0001 undersize. Oversize is not allowed.

In other words, on a lathe, if there is to be a deviation from perfect flatness ( which is impossible) it must be concave, not convex.

How To Run a Lathe is old, by South Bend, same age as my lathe.

[This message has been edited by Evan (edited 12-18-2003).]

12-18-2003, 03:25 PM
DR, if you will look at my first post, you will see that I quoted the spec sheet for Sheldon lathes as being concave from 0.000 to 0.0008" on a 10" diameter. That is pretty hard to ignore.
I would guess that the more precision the machine, the smaller the number would be, but there has to be a tolerance to anything.
Evan, I don't see how the angle of the tool is going to change as it moves across the face of a cut. If it is 90* to the work 5" from center, how can it not be 90* at any other point? I can see how it would change in OD turning, but not a facing operation.
I use the same tool to face something 6" in diameter as 1/2" in diameter. My lathe has a VFD, and I can monitor the current while cutting. If I take a heavy enough cut to see an increase in current, it will decrease as the tool approaches center. Thus the assumption that forces are decreasing.
At any rate, the measurement appears to be taken with test bar or parallel rather than by means of a test cut.

Al Messer
12-18-2003, 04:07 PM
Gee, Fellers, I didn't mean to start WWIII, but are today's lathes that much more accurate that the Clausings & SB's we trained on back in the '60's? Our instructor kept harping that "it was the man not the machine that made the difference" in the accuracy of a piece of work.

When I was in school, items of this nature would routinely be ground on the surface grinder after machining, but since I don't have one of those either, I was hoping for a faster method to produce them than on the Shaper.

C. Tate
12-18-2003, 04:20 PM
Face it. You are not going to get something "flat" on a lathe or a mill. To get it flat you will have to surface grind it. You may be able to get it flat enough but "enough" is a realative measurement. How are you guys measuring the convex or concave form on the face of these parts.

12-18-2003, 05:12 PM
Dr and others:

It sounds like we're talking 2 separate issues here. First, I don't believe the lathes are built with an angle other than a perfect (plus tolerances of course) 90 degrees to the spindle axis. Again, see the machine reconditioning book.

Secondly, the material, cutting tool, setup, etc. will all impact the final result.

Is it possible that over the years, the lathe has been blamed for the effects of these other factors? http://bbs.homeshopmachinist.net//wink.gif


[This message has been edited by nheng (edited 12-18-2003).]

12-18-2003, 08:23 PM
I didn't pull the concavity numbers out of thin air. If you don't believe the numbers then take a look at the test sheets of a Monarch 10EE, one of the finest, if not the finest lathe ever made, or any other manufacturers test sheets. Every test sheet I have ever seen states the same tolerances, or better.
Harry Bloom

12-18-2003, 08:59 PM
Harry: I just checked a copy of an actual inspection document for an Emco Super 11 and it shows the angle formed by a faceplate to the axis of the spindle in sort of a geometric tolerancing method. The faceplate falls within a 0.0005" tolerance zone (two parallel planes 0.0005" apart).

The objective appears to be zero concavity and the lathe tolerance will depend on the precision of the machine.

Perhaps there was a time when older machines did have an intentional angle to the cross slide travel ?? But not now.


12-18-2003, 09:01 PM

The faceplate may fall there (I would expect it) but what is the relationship of cross slide travel to the face plate?

[This message has been edited by Evan (edited 12-18-2003).]

12-18-2003, 09:30 PM
[QUOTE]Originally posted by DR:
Until further proof "

MACHINE TOOL RECONDITIONING and Applications of Hand Scraping by Edward F. Connelly.
Published 1955 by Machine Tool Publications,St. Paul, USA
SECTION 26.65 "Final Alignment of the Carriage Ways", Page 304 thru Page 306.
Recommended Standards, to face hollow or concave only on 12 inch diameter.
Tool Room Lathes 0 to .0005 inch
12 to 18 inch engine lathes 0 to .001 inch.
20 to 36 inch engine lathes 0 to .001 inch.

If you can get a copy of the book, he explains exactly how the test is done to determine this tolerance.

12-18-2003, 09:33 PM
I do not think it is an intentional angle, but a method of stating tolerances so that a convex surface will not be produced as that is less desireable than a slightly concave.

Some of the other specs show the headstock to be pointed up hill 0.0-0.0005" in 12", the tailstock the same. Vertical alignment of headstock and tailstock, the tailstock is to be 0.0005"-0.001" high.

The brochure is dated 1/3/84.

12-19-2003, 01:29 AM
Evan: I asked myself that after posting and re-reading the inspection sheet. The machine recon. book pretty much specs the cross slide at 90 degrees, tested with a parallel held in a 4 jaw chuck and pre-adjusted square by reading with a dti and flipping it over and back. The "hollow" or "concave" on 12 inch only which mac1 mentions is for chuck held work (I think but I need to read it again too) while the cross slide spec preceeds it by several pages.

12-19-2003, 11:37 AM
The "hollow" or "concave" on 12 inch only which mac1 mentions is for chuck held work (I think but I need to read it again too) while the cross slide spec preceeds it by several pages.

On page 305,first paragraph on upper right he says "Since a unilateral tolerance is allowed as shown Fig. 26.64, the dial can have a minus reading only when moving from the front side of the lathe towards the center. Conversely, it can have a plus reading only, when moving away from the center, towards the front side of the lathe"
It sounds like he is saying it is indicating concave.

12-19-2003, 12:11 PM
Mac1: It sure sounds concave, if anything. The unilateral tolerance definitely does not permit a convex face. I usually need to read this stuff two or three times before I get it. Maybe I should read it two or three times before I post too http://bbs.homeshopmachinist.net//wink.gif

It still is a tolerance of course, and the objective is zero, slightly concave if non-zero.

Al Messer
12-19-2003, 01:38 PM
So, are mould blocks with slightly concave faces O.K. for pouring molten metal---safely?

12-19-2003, 02:18 PM
Well, why didn't you say that?
You would probably never see the concavity produced on something of this size.
I am guessing something like a bullet mold here. If it is larger, numbers will change, but math won't.
Assume mold block is 2" square (it will be smaller), and concavity is the maximum of 0.0008"/10 diameter (it would be less), the maximum concavity on a 2" square would be 0.00016". If the mold is 1" square, and concavity is mid range, the concavity produced would be 0.00004".
If that is a concern, it could be easily removed by lapping on a flat plate.
Now we can have a discussion of what a "flat" plate is, and how to produce one, and after that, on the best way to lap a surface.
Isn't this fun?

[This message has been edited by JCHannum (edited 12-19-2003).]

12-20-2003, 04:44 AM
FWIW, I just dug out the factory test certificate for my Chinese 13x40 lathe (yeah, I know!). It gives the facing error as 0.015mm concave in six inches diameter.


12-20-2003, 01:21 PM

Speaking of factory test certificates.... your post reminded me of something.

A few years back I bought the remaining contents of a machine tool dealer. Included were three smallish wooden shipping crates, each contained a complete taper attachment for a Jet lathe. I noticed the inspection certicates and was impressed with their thoroughness in documenting the accuracy of these parts in what seemed like every possible way.

Out of curiosity I compared the certicates from all the crates. What do you know? They were all exactly the same, right down to the little ink smudges from someone's thumb!!! A less trusting soul than myself might suspect some funny business with the inspection certicates.

12-20-2003, 11:24 PM

If I had a nasty, suspicious nature (which I'm sure I don't), I would have wondered at the consistency of shape and size of the Inspector's ticks on the test certificate - it almost looks as if they were pre-printed on the certificate, but that wouldn't be the case, would it?

However, to be fair, I've had a good run from the machine so far, and it is accurate enough for what I do.


[This message has been edited by franco (edited 12-20-2003).]

12-20-2003, 11:42 PM
Double posted.....deleted

[This message has been edited by spope14 (edited 12-21-2003).]

12-20-2003, 11:45 PM
My thoughts on this one. Basic machine rigidness issues. NO MATTER THE MACHINE, SIZE OF THE MACHINE, OR THE TOOLS USED, PHYSICS TAKE OVER.

Rigidness - the facing tool at center of the part - especially a ten or twelve inch diameter part. The facing tool at part center / X axis "0", or on part exact center is actually at the strongest part of the set-up, or better said, the cross slide/ and compound are much more centered on and over the carriage, and also the tool is much more centered on the bedways of the lathe - probably exact center of the ways. Basically said, all conditions of rigidness being at maximum condition are satisfied.

The further out from centerline of the part, the tool is now traveling "Off the ways", and to the outer extremities of the carriage, or to look at a diving board, the center of the part is the pivot, and now the tool point in facing is away from the pivot point, thus pivots "OUT" from the head stock, thus causing the outer edges of a part to be the "high points" by the material and the physics of cutting pressure pushing the tool away from the work.

When facing "In" the tool will actually leave a slight "start ring" on a stock surface OD before it pops back - very un noticable, unless one like myself sees this many times and wonders "what the heck?????" The tool pushes away, but getting to center, (I lock my carriage facing), it settles back to where it should be.

[This message has been edited by spope14 (edited 12-20-2003).]

Al Messer
12-21-2003, 07:11 PM
Well, I have learned a bit from this experience. The cubes that were clamped near the edge of the rim of the faceplate were not as concave as the two in the center. Boy, were they ever off from being square and flat. I solved the problem by finishing them off in the Shaper even though it did seem to take forever.

12-21-2003, 07:47 PM
Al, you might want to look at your faceplate.
It could pay to take a skim cut on that first to be sure it is "flat".

[This message has been edited by JCHannum (edited 12-21-2003).]

12-21-2003, 10:17 PM
Ah may have taken forever, but remember, great work takes time, and pride lasts forever.

Great thread you started here, I have learned quite a bit, and had to experiment and look over specs before I answered the thread......

Al Messer
12-21-2003, 10:30 PM
JC, I did skim it shortly before trying this job. I leaned a long time ago that, sometimes, faceplates will warp on you, especially if they are still "green". I still sometimes wonder how much of this discipline is Science and how much is Art.

12-22-2003, 02:00 AM
...or maybe the magic "smoke" escaped! http://bbs.homeshopmachinist.net//biggrin.gif

I hate it when that happens - I blame the gremlins.

12-23-2003, 06:01 PM
hell.....make 'em as square as you can on your faceplate and true 'em up with a file and some lapping compound on a surface plate (or a pane of glass)....man, hate these discussions about tolerances of .00005......lolol....Merry Christmas All !!!