Facing rectangular stock in a lathe

Greg,
I'm curious. Did you face it from the center out, or from the outside in. If from the outside in, that sound like it could cause what you describe. As to the spiral pattern, could be related to the feed rate (too fast?).
I'll be interested in other responses. I'm new to this as well (with the same lathe).
-Bruno
Greg Menke wrote:

I think you may be correct in your suspicions. Let me assume you are using a high-speed steel tool. Check the cutting edge on your tool for: 1. Sharpness 2. relief 3. rake 4. Position
1. Feel your edge. It should be sharp enough to cut skin, though perhaps not as sharp as a razor blade.
2. Look at your edge. The side and end should angle away from the workpiece so that only the cutting edge makes contact. If the side and top do not make a clean, sharp angle, your tool is either worn or ground improperly. 3. Look at your edge. The top may be flat or sloped down from the cutting edge so that the tool peels off a chip like a snow plow picks up snow and throws it aside.
4. Look at your setup. The tip of the tool should contact the work piece at an angle so that the side of the tool is nearly -- but not wuite -- parallel to the surface to be faced. Height of the tool should set to bring it to the exact center of the work piece.
You get best results when everything -- angles, speeds, fluids, sharpness, etc are set exactly right, but you should get acceptible results with a tool that is kinda sharp, has some relief, some rake, and is set to contact the work properly.
Pat

[ ... ]
I think that you may be experiencing rocking of the compound, or give in the lantern style toolpost (which I believe is all that came with the machines). To minimize the rocking of the compound while facing the stock set it parallel to the axis, and tighten the gibs as much as practical, since you should not need to be feeding with that. (For parting off, you want the compound at right angles to the spindle's axis, so you are stuck having to change it from time to time -- aside from setting up the proper angle for thread chasing.
This is a relatively weak machine, so you have to do more to take advantage of every boost to rigidity that you can.
A quick-change toolpost would improve the rigidity -- but the smallest standard one (Aloris AXA series, Phase-II series 100) is too big -- though there are other designs which work well on a machine of that size -- such as the one on my Emco-Maier Compact-5/CNC. The same style of toolpost is made for Myfords in England -- but it will probably be more expensive than the relatively high-production Aloris clones. I usually advise the wedge style with Aloris clones, but you can probably get away with a piston style if you can find one to fit, given the lack of rigidity of the rest of the machine. (I have one of these -- which I don't use much these days, so I know just how flimsy it can be. Is yours the later style with roller bearings, or the older one with split plain bearings? Wear in the latter can add more flex to the setup.
Of course, the grinding of the tool bit can make a big difference, too, as you already mentioned.
Good Luck, DoN.

See the line from MASH where Clinger said his parents put a violin in the crib, hoping to make him a musician? He replied something to the effect "what did I know from violins - I started sucking on the frets...".
Clinger knew more about violins as a child than I do about what I am about to tell you ;-) so take it for what it is worth.
On a MILL (what's a lathe?), I read a given cutter NEEDS a certain cutting pressure (feed rate?) to cut or it will just generate heat and self-destruct. Balance this with a need that the given pressure must not cause deflection of the work piece for an accurate cut, and the solution becomes either reduce deflection (make your "setup" more rigid if the deflection is not the workpiece itself), or you *must* use a cutter with less surface area requiring less pressure/feed rate.
It's been said two or three times in a matter of offering a (correct) solution - but I offer this a the reason for the problem's existence (sure hope I ain't steering you askew!).
Basically what I hope I said correctly is, you GOTTA push to cut - and smaller cutting surface areas require less pushing (causing less problems concerning deflection, resonance, chatter, or ?).
I probably should have hit the delete key at this point so please re-read my above "disclaimer". I better re-enter lurk mode...
Best,
SBK
Greg Menke wrote:

I was using the compound set to 45 degrees, positioned with the toolpost pretty much right on top of the crossfeed. I used a round-nosed tool, applied straight to the face of the work- the various bit angles looked reasonable, but what do I know... I locked the carriage by engaging the feed lever with the feed itself disengaged. Next time, I'll try leaving the compound parallel to the crossfeed.
Its a timkin bearing model lathe, but not too heavily used. OTOH, its a pretty old lathe and light too, so even light use adds up I suppose. I know exactly what you mean about it not being very rigid, my dad has a 9" South Bend and its like a rock in comparision.
Thanks,
Gregm

I tried the cuts in both directions at a variety of different feed rates and depths. Even light couple mil cuts taken slowly didn't obviously improve the finish over a 20 mil hogging cut. Looking at the faces, aside from the "smears", there isn't much difference in the spiral pattern from the "inside" portion of the face to the portion out on the corners where the tool hops from edge to edge- there I took things slowly to keep the battering down.
In this case the spiral pattern is pleasing to look at, though I'll be trying to polish it out a bit (I left ~5 mils over for sanding) and theres no need for anything like a factory finish.
Thanks,
Gregm

All the facotrs you mention play a very important part in obtaining a nice finish. Just take one of those factors out and it leads to a poor surface, so I would try honing your skills on sharpening toolbits, or buy a brazed tip carbide cutting tool and give it a go, and see if things improve. 12L14 should turn very nice. To minimize deflection secure the cutting tool in the toolholder leaving as little of the exposed cutting surface as possible . Less overhang equals less deflection, and better finish.
Visit my website: expressed are those of my wifes, I had no input whatsoever. Remove "nospam" from email addy.

Not the best - snug down the gibs and turn the compound parallel to the ways.
Problem two. Change over to a tool with less surface contact with your workpiece - a sharp facing tool with a slight radius on the edge.
That's the third problem - lock the carriage down with the lock screw - there's too much slop in the feed setup to accurately locate the carriage.
In spite of what DoN said, I've always had better luck with setting it parallel to the ways - and using the compound as the infeed rather than the carriage handwheel. Be sure the compound gibs are snug.
If the machine is setup well it will be as good as a 9" SB. But most novices have the compound and cross-slide gibs a bit on the loose side.
The problem you see on the face of your part is caused by the interrupted cut you are taking. The tool is entering and leaving the workpiece four times each revolution, until you are inside an inscribed circle in the square.
Interrupted cuts are almost as tough as knurling to do right, and small machines are flexible enough to allow the vibration from an interrupted cut to vary the depth of cut, and allow a pattern like you are seeing, to develop. Even rigid lathes will show that 'inscribed circle' effect when facing square stock.
So basically you want to:
a) use a sharp tool with plenty of relief. Face from outside inwards.
b) use a tool that minimizes surface area contant with the workpiece - ie no round nose or form tools.
c) keep the speed and feed low to mimize excitation of of the vibration modes in the system.
d) make the setup as rigid as possible:
i) lock down the carriage with the lock bolt ii) snug down the gibs iii) use the most rigid toolholder setup possible
If you don't have a QC toolpost, and are using a lantern type toolpost, don't despair.
Toss out the armstrong tool holder and obtain one of the largest HSS tools that will fit inside the lantern toolpost. Grind tool appropriately. You will need to use another toolbit, or a square bit of CRS as a shim, but get the cutting edge of the HSS tool to the correct center height. Keep the overhang of the tool to the bare minimum.
Mounting the HSS tool right in the lantern toolpost and getting rid of the toolholder does wonders for eliminating flex on smaller machines.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Greg Menke wrote in news: snipped-for-privacy@europa.pienet:
Deflection is your problem, probably in both the tool and workpiece.
When making interrupted cuts, the name of the game is RIDGITY. You want the end of the part as close to the chuck jaws as possible, and you want the stiffest tool you have, with the least hangover of the toolpost as possible.

Hey Greg,
Well, nobody else seems to have mentioned it, but you might try a left hand cutting tool which has the major cutting part nearly parallel to the shank, and the radius from that side to the front edge. The left hand edge of the tool would be about 3/8 inch "long" on mine. Start at the centre and cut outwards. Set at a very slight angle so the radius is just trailing, it "shaves" like a razor across a wide surface, rather than cutting in a sense like a knife or chisel. Sort of flat faced, if you will.
Take care.
Brian Lawson, Bothwell, Ontario. XXXXXXXXXXXXXXXXXXXXXX

Greater contact area requiring more force to apply the tool, thus more deflection?
By "locate" do you mean rigidly fixed in position? I was using the feed screw simply to hold the carriage in position, then used the compound to apply the tool and the crossfeed to draw the tool across the face. I forgot about the lock screw... oops.
I worked over the gibs before starting, they're all pretty snug, not so tight that you need leverage to turn the cranks, but enough so you need to apply a little force.
I've seen ads for them in Home Shop Machinist & others, they look nice- are there particular considerations that apply for fitting them to a little Atlas?
The bearings in my live center are in nasty shape, so I need to do a little shopping fairly soon anyway...
Cool idea. I'll give that a try next time.
Thanks,
Gregm

Exactly. Tools with large contact area, like threading, form, or cut-off, put extreme requirements on the rigidity of setup.
It's quite probable that some of the flex that contributed to the 'inscribed circle' was simply the carriage moving back under the strain of the cut. Lock it down with the lock screw to eliminate that.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Greg Menke snipped-for-privacy@toadmail.com
wrt flat facing and as to lathe
It's hard to post when you're speaking of both technique and machine quality.
Southbend lathes are/were a general line product, nothing more. Facing results won't be determined by the differences in machine.
Are you facing under power only? If you are, trying hand techiques will advance the possibilities. And why not center the work first? Whether or not in a four-jaw, I think that would help the assessment.
Frank Morrison

No! Not parallel to the crossfeed, but parallel to the lathe spindle's axis -- and the bed ways. The setting at right angles to the spindle axis and ways is for *parting* cuts where the stresses are different, not for facing cuts.
As others have already pointed out, the round-nosed tool is increasing your cutting forces.
And "locking" the carriage with the half-nuts is not very rigid at all. There is some backlash in the half-nuts, and often more in the end bearing of the leadscrew. The carriage lock is what should be used for that purpose -- a square projection to the right of the cross-slide, IIRC. It should fit the same wrench which is used for tightening the lantern style toolpost.
Then the spindle should be fairly rigid, at least.
Part of the problem is just the difference in size. Part of it is the square ways, instead of the inverted 'V' ways which give better overall support to the machine. Part of it is the way the compound attaches to the cross-slide, with radial setscrews trying to pull it down using a round dovetail projection from the cross-slide, instead of a cylindrical projection or hole, and vertical screws through the compound's base into the cross-slide for wider support.
Good Luck, DoN.

If you'll look back at my article (or even at what he quoted), you will see that I suggested the same orientation as you -- and had the other orientation (parallel to the crossfeed) suggested for parting, in reference to another recent thread.
Enjoy, DoN.

Except that he was using the longitudinal *feed* to attempt to stabilize the carriage. There's even more uncertainty in that than in the half nuts/lead screw combination.
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

Yep, I guess I had my brain stuck on the parting off thread. Sorry for the confusion!
Jim
================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================