PDA

View Full Version : Unwanted taper



Stu
03-03-2015, 04:54 PM
My lathe is cutting an unwanted taper. A little backround first, it's a Craftsman/Atlas 12x36 I just bought. It has 1/2" thick flat ways that have .002" wear on the outboard way for about the first 8" from the head stock, everywhere else measures spot on. I don't have a machinists level, so I got it close with a regular one and used "Rollies Dads Method" http://igor.chudov.com/manuals/Rollies-Dads-Method-of-Lathe-Alignment.pdf
to align the lathe. Please advise me if I did this correctly, I chucked up a steel bar 16" long with the same dimension through out, mounted a dial indicator in the tool post made sure it was on center and found the high and low spots next to the chuck. I then zeroed the DI on the average. I moved the carriage to the end of the bar and found the average of the high and low and compared that to zero. The difference was .007, I then shimmed under the rear of the lathe at the HS end till the two averages met. I then repeated the whole process and found a difference of .001, I shimmed again until they matched. Now it is still cutting a taper of about .003 in just 2". My test piece is aluminum 2" in diameter, I am taking .005 cuts and it measures smaller next to the chuck. There is no provision to adjust the headstock as it fits snugly between the ways. Any advice?

Stu

2ManyHobbies
03-03-2015, 05:27 PM
I'm not sure if you are turning between centers or not. If so, most lathes let you adjust the tailstock and it sounds like the tailstock needs to move towards you just a tad.

If you aren't turning between centers, then my input probably won't be that useful.

Stu
03-03-2015, 06:25 PM
I'm no turning between centers, the test piece is 4" long

Bill

boslab
03-03-2015, 06:25 PM
LH/RH shimming sounds off whack somehow, ie front and back as it sounds like a sort of twist thing, does that make any sense, I've only ever done a small myford so experience limited, you clocked horizontally but how's axialy, ie 12 o clock and 3 o clock?
If 3 o clock is out then as has been said your tailstock is offset by a few thou
Turn a collar both ends the same, set the crosslide to zero, measure the diameters and move the tail til they match
Mark

SGW
03-03-2015, 08:01 PM
That sounds correct to me, I think. Just to clarify: you got the indicator to center its movement on zero up near the headstock, for example in one revolution of the test bar the indicator needle might go -3, +3.
You then moved the indicator out to the end of the bar, but did not touch its adjustment. At the end of the bar, the reading range might be +1, +7. By shimming one leg of the lathe, and only by shimming the leg of the lathe, without touching the indicator, you got the indicator to read -3, +3 in one revolution of the test bar.

Is that what you did?

Stu
03-03-2015, 08:27 PM
That sounds correct to me, I think. Just to clarify: you got the indicator to center its movement on zero up near the headstock, for example in one revolution of the test bar the indicator needle might go -3, +3.
You then moved the indicator out to the end of the bar, but did not touch its adjustment. At the end of the bar, the reading range might be +1, +7. By shimming one leg of the lathe, and only by shimming the leg of the lathe, without touching the indicator, you got the indicator to read -3, +3 in one revolution of the test bar.

Is that what you did?

Yes, exactly

Stu

Forestgnome
03-03-2015, 08:35 PM
Did you measure the diameters at both ends of the bar?

Stu
03-03-2015, 09:23 PM
Did you measure the diameters at both ends of the bar?

Yes, both ends are the same diameter.

Stu

Don Young
03-03-2015, 10:13 PM
I suspect headstock alignment is off a bit. Often there is enough clearance to rotate it slightly.

TGTool
03-03-2015, 10:47 PM
I've also got a Craftsman 12 x 36 and did Rollie's method first. I still felt like it wasn't quite right so used a level and it showed I should do some additional (or different) twist adjustment. Adjusting it vial the (precision) level seemed to fix it up. I've never quite understood it because in theory the Rollie method should be fine, but this was just my experience.

LKeithR
03-03-2015, 11:52 PM
Yes, both ends are the same diameter...

If the piece is the same dimension on both ends there is no taper--sounds like you've got it dead nuts...

darryl
03-04-2015, 02:33 AM
I see two scenarios possible here- one is to chuck a fairly large diameter test piece and take light finishing cuts with a sharp tool- the other is to use a small diameter test piece with a tool that may not be sharp, or ground with proper relief angles. It's very easy to deflect a work piece at the unsupported end, thus leaving that end large compared to a spot closer to the chuck jaws. You do need to know if this is happening (and it's a fairly common thing).

As soon as you start using a tailstock center things will change, and it's not always obvious what is happening when measurements show inconsistencies.

Gary Paine
03-04-2015, 03:26 AM
You shouldn't be expecting much better with 4 inches sticking out of the chuck and no tailstock support. Drill the end and install a center and the taper will possibly go away.
A chuck with worn jaws will allow a little flex of the work away from the cutter. Up next to the chuck you should not see it.

Stu
03-04-2015, 06:49 AM
You shouldn't be expecting much better with 4 inches sticking out of the chuck and no tailstock support. Drill the end and install a center and the taper will possibly go away.
A chuck with worn jaws will allow a little flex of the work away from the cutter. Up next to the chuck you should not see it.

The chuck does have some bell mouthing on the jaws, I'll check that out, Thanks

Bill

Stu
03-04-2015, 06:52 AM
I see two scenarios possible here- one is to chuck a fairly large diameter test piece and take light finishing cuts with a sharp tool- the other is to use a small diameter test piece with a tool that may not be sharp, or ground with proper relief angles. It's very easy to deflect a work piece at the unsupported end, thus leaving that end large compared to a spot closer to the chuck jaws. You do need to know if this is happening (and it's a fairly common thing).

As soon as you start using a tailstock center things will change, and it's not always obvious what is happening when measurements show inconsistencies.

Good points, I'll try that. Thanks

Bill

Norman Bain
03-04-2015, 02:54 PM
I cringe every time I see the "Rollie's Dad's" method of lathe alignment mentioned.

JMHO: but there are way way better ways of establishing the ways are aligned to the headstock.

Cheers,
Norman

Black_Moons
03-04-2015, 03:04 PM
I've also got a Craftsman 12 x 36 and did Rollie's method first. I still felt like it wasn't quite right so used a level and it showed I should do some additional (or different) twist adjustment. Adjusting it vial the (precision) level seemed to fix it up. I've never quite understood it because in theory the Rollie method should be fine, but this was just my experience.

I untwisted my lathe by turning a shaft and adjusting lathe feet till it turned straight... after using a level to 'get it in the ballpark'

Stu
03-04-2015, 05:39 PM
I untwisted my lathe by turning a shaft and adjusting lathe feet till it turned straight... after using a level to 'get it in the ballpark'


I started to do that way and was getting closer, but couldn't get it quite right, then I read about the Rollie method. I think maybe I'll go back and try it that way, I think I just have to be more patient and go in smaller steps.

Bill

lakeside53
03-04-2015, 07:34 PM
You guys must all have rubber lathes especially if you don't have the combination hold-downs (into the floor) and jack-up screws. I have 8 adjusters (4 per pedestal). If I turn one a 16th of turn, another simply rises off floor. Gravity on 3500lb cannot twist or untwist my sturdy lathe in any material way I can measure it, and I have a zillion $$$ machinist level.

If I had hold-downs under each of the adjusters into 12 inches of concrete/rebar (factory recommendation) , maybe I could exert reasonable force in combination with the jacking screws, but...

Even with smaller lathes and 4 points, without hold-downs in concert with jack screws, you're going to find on corner "light' and most weight on three.

Emco 10 and 1 inch lathes mount the bed at two points only - one to each pedestal. You can jack around with the feet day long and you won't "twist" the bed.

Black_Moons
03-04-2015, 10:57 PM
You guys must all have rubber lathes especially if you don't have the combination hold-downs (into the floor) and jack-up screws. I have 8 adjusters (4 per pedestal). If I turn one a 16th of turn, another simply rises off floor. Gravity on 3500lb cannot twist or untwist my sturdy lathe in any material way I can measure it, and I have a zillion $$$ machinist level.

If I had hold-downs under each of the adjusters into 12 inches of concrete/rebar (factory recommendation) , maybe I could exert reasonable force in combination with the jacking screws, but...

Even with smaller lathes and 4 points, without hold-downs in concert with jack screws, you're going to find on corner "light' and most weight on three.

Emco 10 and 1 inch lathes mount the bed at two points only - one to each pedestal. You can jack around with the feet day long and you won't "twist" the bed.

Nah, I definitely took the twist out of the bed by just changing the feet. My lathe bed also has 4 mounting points to the two cabinets and very little to connect the two cabinets.

Old Time
03-04-2015, 11:27 PM
Just an observation, my Logan would turn a taper of between .003 to .006 in a three inch length cut, it either tapered larger by the chuck, then the next time it was larger near the tailstock. I bought a precision level, it was level, then I tried moving the head stock, no difference. I checked the bed for wear, about .0002. After a good cry, I figured it was just the nature of the beast. Then one day I was turning something and I noticed the compound appeared to move. The problem was the gibs on the cross slide were just loose enough the the cut would take up the slack. I adjusted the gibs and drilled and tapped another screw hole in the cross slide with a brass tipped locking screw. With the cross slide locked it cuts straight, .0002 over a 6 inch cut. Check for play in the cross slide and compound gibs.
Old Time

darryl
03-05-2015, 12:32 AM
It's probably fair to say that there isn't just one cause. That would be too simple. I'll suggest my first idea again- to turn a test piece from a large enough diameter bar that flex shouldn't be a problem. With the lightest of finishing cuts you might avoid problems at the chuck itself if the jaws are bellmouthed. You should be able to discount that as a source of the problem.

Using a tailstock center could easily introduce an error- if the center is not measurably 'on axis', it will flex the end of the bar in some direction. The result of this is that you may be fooled into believing that the headstock is out of alignment- the lathe may be turning a taper, but that's all you know at this point. The headstock could be true to the ways and you could still find that it's turning a taper.

Wear on the ways is probably the worst situation as it will change the position of the cutting tool as the carriage goes left/right. The change won't be linear, but the further from the chuck the carriage is, the more likely it is that the carriage tracks the ideal path. Of course this isn't practical for turning a test bar as you would need so much 'stick-out' to have the carriage riding over an unworn spot that you couldn't avoid having the workpiece flex.

About the only thing you could know for sure is if you were to be able to make a divot on the end of a test bar without throwing the bar into an eccentric motion, the divot would be inline with the spindle axis. That's not to say that axis is parallel to the bed, but it could be a useful indication of whether the chuck is capable of holding a work piece without it going eccentric as machining forces act on it. You would turn a section near the outboard end, then make the dimple, then check to see if the section you turned got into an eccentric motion. If it didn't, you might reasonably surmise that the chuck is holding the work piece adequately, and thus you could eliminate the chuck in your quest to find where the taper is coming from.

Here's a test- probably useful- chuck a short piece, face it and create a dimple in the end. Retract the ram in the tail stock, snug the quill, and mount a center. Put a dab of grease in the dimple and tuck a small bearing ball into it. Get a flat disc or square of material and lightly punch a mark roughly into the center of it. Bring the center up to the bearing ball and with the mark in the disc nested onto the center, crank the ram just enough to pinch the disc to the ball. The disc will likely lean one way or another, and you would adjust the tail stock in such a way that the lean disappears. Since the tail stock is not likely to have worn it's ways to any extent, it's reasonable to assume that it can traverse the ways without going off kilter like the carriage is prone to do if riding worn spots on its own ways. Now mount a longer work piece and create a dimple in the end of that. Do the same test with the disc and see how the disc leans now. This is likely to be one of the better indications of whether the head stock is aligned with the ways.

For this test to work the best, you would turn a point on the end of the test bar instead of a dimple. A disc sandwiched between sharp points will give a more amplified indication of whether the points are aligned. A disc that leans towards the head stock at the top indicates that the point on the test bar is low- a disc where the front edge (closest to you) leans towards the head stock indicates that the test bar point is behind the tail stock point, etc.

You are fairly likely to find that when the height of the tail stock center is aligned, it will remain in close alignment for test bars of different lengths. It's more likely for a head stock to be out of alignment in the horizontal plane- in other words the test bar point will closer to you or further from you at different distances from the chuck. This will tell you if the head stock needs alignment.

Throw a twisted bed into the picture and this test is largely out the window due to the tail stock corkscrewing as you slide it along its ways.

All in all it's complicated-