Question on lathe accuracy

I'd guess it was a function of how the two bars were chucked. The external cut was on a tailstock-supported bar (yes?) The internal cut - I'm guessing here - wasn't supported by the tailstock? ;-) The tailstock itself could be *marginally* offset. This can be best checked by turning something between centres (i.e., no chuck).

Any which way, your errors are trivial (IMHO) and could be attributed easily to just the way you held your mouth when you tightened up the chuck.

Make sense?

-- Jeff R.

Sounds like workpiece deflection. At 2", the piece may not be bending, but it may be moving in the chuck jaws.

John Martin

The errors you see are in the correct direction for workpiece spring. For the OD cut, the part moves away from the tool so it will have the largest diameter far away from the headstock. Here you can imagine the part moving away from the tool, springing towards the back of the machine.

For the ID cut, the part would be springing towards the front of the machine, because that is how the force is being applied. So it will bore small at the very end of the part.

You could re-try this test by cutting some aluminum, using a very sharp tool, and taking only a thou off on the last cut.

Also note (as others have mentioned) the chuck may be flexing and allowing the workpiece to move, or the spindle bearings may have enough slack in them to cause a similar effect. You can check that with a good tenths reading dial indicator.

Jim

================================================== please reply to: JRR(zero) at yktvmv (dot) vnet (dot) ibm (dot) com ==================================================

x??TËnÛ0¼ë+ö??á¦=m i{p=ä(j-±¦¸*Qô÷¥å$BzéÍ?³3ÃÙm?&fè?Ð%ó§æÆÝ^õù3h0L&ø?îåé(±#oòÀ{z?Zî(

the bed is worn most at about 8 inches from the chuck?

Was the bar supported at the outboard end? If not, you'll get deflection when cutting stuff with that much overhang. If you did, you might want to check your tailstock centering with a couple of dead centers.

On your boring project, a lot depends on how rigid your setup is, if you're using a boring bar a lot smaller than your 2" hole diameter, you'll get deflection that way, too. I usually make a number of finish passes at the last cross slide setting to compensate for bar deflection. I also try to make the boring bar a fairly close match to the hole size if I'm trying for a good, parallel hole. A lot depends on how deep you're boring your hole and the amount of boring bar that sticks out past the holder.

You might also want to check to see that your lathe bed isn't twisted. If your Boxford is the copy of the 9-10" Southbend that I think it is, they aren't that rigid. They're nice small lathes, I've done a lot of work on a 9" Southbend, but they do flex and if your shims have slipped and the bed has a little wind in it, you'll end up with some untrue results.

Stan

Wear on a lathe bed usually occurs towards the headstock end of the bed, but not right at the headstock. It occurs where the carriage moves the most over time. You end up with a low spot in the ways. It is easily detected by putting a dial indicator on the carriage and reading the bed where the tailstock slides, which usually doesn't wear much. As the carriage goes lower sliding in the worn area, the tailstock surface will appear to rise.

Such wear caused the tool to move vertically. Vertical movement of the cutting tool is treacherous. It will affect small diameters more than large diameters. You cannot correct for the effect by offsetting the tailstock. Cutting on the outside of a bar, the cutting tool cuts lower away from the headstock, in the worn area. This has the effect of making the cut larger in diameter away from the headstock. If the tool is set to bore a hole, and is on center line at the end of the workpiece which corresponds to the low spot in the bed, it will rise as it goes toward the headstock, causing it to cut a larger hole near the headstock.

This is a common wear pattern on old lathes. The only really good way to fix it is to straighten the bed. Brownnsharp

Thanks for your help and advice. A few follow up notes:

I did not use a tailstock centre when turning the two inch bar. The last couple of cuts were 0.001 so I thought that this would keep deflection to a mimimum.

The bed is worn in the area close to the chuck, 4 or 5 inches along the bed from the chuck.

The bed is mounted on jacking screws so that I can twist the bed and compensate for some of the wear. But it's a bit of a compromise.

I was happy with the results I got when turning the outside diameter but I was very supprised with the results from the internal turning.

I used a 0.75 diameter boring bar for the internal cut and the last couple of cuts were 0.001 to try and keep deflection to a minimum.

The headstock should be in perfect alignment because it mounts on the bed ways, there are no adjustments. The taper bearing are in good condition and correctly adjusted.

The lathe has seen a lot of heavy work over the years and I am probably expecting too much from this "small lathe" (similar to a Southbend).

Regards, Martin Perry

Trolling for info here...

If your bed isn't really hell-ass worn, like, visibly (>1/32" below head/tailstock levels), which if it were I can't imagine things would be sliding all to well anymore, then on say a 1" dia. work piece, you'll have damn near nil change in diameter. Even by Harold's standards.

Let's see. Say it's down 10 thou. A right triangle is drawn, one leg from the center of the work, out straight (where the tool should be) and the other down from here (to where the cutter *is*). The straight leg is the radius (.5"). The 10 thou drop represents the other leg, with the hypotenuse representing the new work diameter. So we have D' = sqrt(D^2+h^2), which here comes to.... nearly a tenth. For a small bit of work, say 1/8" diameter and 20 thou wear (ouch), this comes to 3.1 thou, about 2.5%. Depending on what you're doing this could be significant.

But to show any sign of this difference (if you're measuring as you go, that is) you need a workpiece longer than the "step" of wear, and 1/8" anything is notoriously flexible, so you'll need a follow rest, which combined with the flexibility of the work will utterly negate any way variations...LOL!

Tim

-- "I have misplaced my pants." - Homer Simpson | Electronics,

- - - - - - - - - - - - - - - - - - - - - - --+ Metalcasting and Games:

Just to put some numbers on it, one thou in diameter is half a thou in radius or, in this case, possible deflection. Your 8" long, 2" diameter steel bar is a cantilever beam.

BeamOut''(30E6,I_tube 2 1)Beam 0 (8 70) pos'n pt. couples shear bend. slope def'l force stress stress 0 -70 -560 -29.71 713 0 0 8 70 0 0 0 0.00009507 0.000507

Only 70 pounds force is needed to deflect the end half a thou. This is pretty small as cutting forces go.

Aluminum has 1/3 the elastic modulus of steel so, as Jim said, the last couple of finishing cuts woulld have to be very light.

Ted

Turning unsupported work 8" long on a light lathe will give you these sorts of results. Mostly, you have probably demonstrated spindle flex. A .001" cut on a nice aluminum alloy with a brutally sharp tool will not need much infeed force. Depending on the workpiece material and the sharpness and profile of the cutting tool, you may or may not get highly accurate cutting on very fine cuts like this. You can also have false edge buildup on the cutting tool and thermal expansion of the workpiece to confound you. Whenever you rough a piece, you need to allow the temperature to return to ambient before making that last finish pass. If the part is still cooling off from the roughing passes when you make the long, slow finish cut, you will also get taper. It is pretty amazing how much a

Jon

Tim is right. I pictured what might happen in my head and was too lazy to run a calculator. And I do appreciate the correction.

A bed that is worn in the middle is not likely to have worn straight down. But I will not look at the "down" vector as particularly important anymore. It is still an easy measurement to determine if you have wear though.

Brownnsharp