What's acceptable for accuracy?

[epanzella]

It's possible to have much better repeatability (which is what I think you want for your test bar) than accuracy.

I know repeatability is essential to what I'm trying to achieve but I don't follow what you're recommending. Where I'm headed is to have a one foot test bar turned on centers with a 3 inch collar near each end and the lathe tuned to make those collars is identical as possible. Is there some repeatability factor that I'm missing? Please bear with me as although I've been making drag car parts for a long time (tie rod, driveshafts, ect) this is the first time I had a machine with which I could even attempt to accurate. On my old machine I could actually see the carriage climb up onto the relatively unused portion of the bed as it got further from the chuck.

 

[george wilson]

You can buy special,precision ground test bars for lathes. Unfortunately,they aren't cheap,and are the type tool you might only use once in a blue moon.

As for drill rod,IIRC,it,being centerless ground,tends to have 5 lobes on it. As said,this can vary,depending upon how the machine was set up. And,it might measure true,as said,unless you have a 3 point micrometer.

Mind you,these lobes are not very pronounced,and are in the tenths or less. It depends upon what you decide is accurate enough for your needs. I am a bit of an accuracy nut,over the top,myself.

 

[John Hasler]

I know repeatability is essential to what I'm trying to achieve but I don't follow what you're recommending. Where I'm headed is to have a one foot test bar turned on centers with a 3 inch collar near each end and the lathe tuned to make those collars is identical as possible. Is there some repeatability factor that I'm missing? Please bear with me as although I've been making drag car parts for a long time (tie rod, driveshafts, ect) this is the first time I had a machine with which I could even attempt to accurate. On my old machine I could actually see the carriage climb up onto the relatively unused portion of the bed as it got further from the chuck.

I just mean that you need to know how much those two collars differ after you turn them, not their absolute diameters. A micrometer with a worn anvil can be repeatable to a tenth but only accurate to a thou.

 

[epanzella]

I just mean that you need to know how much those two collars differ after you turn them, not their absolute diameters. A micrometer with a worn anvil can be repeatable to a tenth but only accurate to a thou.

Oh, yes, I get what you meant now. The bar will start off with 3 inch collars but once I start turning and tuning, I won't care what their absolute OD is as long as they're as identical as I can get them. Thanks for helping.
Ed P

- - - Updated - - -

You can buy special,precision ground test bars for lathes. Unfortunately,they aren't cheap,and are the type tool you might only use once in a blue moon.

As for drill rod,IIRC,it,being centerless ground,tends to have 5 lobes on it. As said,this can vary,depending upon how the machine was set up. And,it might measure true,as said,unless you have a 3 point micrometer.

Mind you,these lobes are not very pronounced,and are in the tenths or less. It depends upon what you decide is accurate enough for your needs. I am a bit of an accuracy nut,over the top,myself.

That must be the boogyman preventing me from getting past .001. I priced a test bar and it was $200 for MT-5. Too pricey for an old retired guy.

 

[Don B]

I've just been reading through this thread and I noticed you've been talking about turning between centers and trying to get as much accuracy as possible, have you used a test bar (unsupported) to check and see if you have any twist in the bedways before you started adjusting your tailstock?
If your ever working with a piece of stock that is long and the center is out you can run it in a steady rest a knife out the existing center.

 

[joshua_]

jelsed the same

 

[epanzella]

I've just been reading through this thread and I noticed you've been talking about turning between centers and trying to get as much accuracy as possible, have you used a test bar (unsupported) to check and see if you have any twist in the bedways before you started adjusting your tailstock?
If your ever working with a piece of stock that is long and the center is out you can run it in a steady rest a knife out the existing center.

When I got the lathe I started by indicating the spindle right inside on the MT5 taper and the movement was so small as to be unreadable. Then I put in the MT5 to 5C sleeve and got something less than .001. (I thought that was a bit sloppy but what do I know?) Then I indicated 1 inch drill rod and got .001 at the spindle and .002 or 3 (don't remember) about 8 inches out at the unsupported end with the rod in a 5C. At the time I couldn't dial in a 4 jaw to less than .002 so didn't bother with that. I adjusted the drill rod in the 5C with the leveling feet until I got it about .001 at the end. (took about two weeks) I've since learned to dial a 4 jaw (thanks to this forum) and can get it well under .001. That is when I turned some junkyard 1 3/4 on centers and saw a taper of .002 on the OD. The tailstock had moved from it's factory setting. This is where I'm at now and want to make damn sure I know what's really happening before I move that tailstock. At the same time I want to make sure I end up with a test bar that I can depend on to align my tailstock in the future. All indicating was done with a .001 x 1.000 inch DTI.

 

[Don B]

When I got the lathe I started by indicating the spindle right inside on the MT5 taper and the movement was so small as to be unreadable. Then I put in the MT5 to 5C sleeve and got something less than .001. (I thought that was a bit sloppy but what do I know?) Then I indicated 1 inch drill rod and got .001 at the spindle and .002 or 3 (don't remember) about 8 inches out at the unsupported end with the rod in a 5C. At the time I couldn't dial in a 4 jaw to less than .002 so didn't bother with that. I adjusted the drill rod in the 5C with the leveling feet until I got it about .001 at the end. (took about two weeks) I've since learned to dial a 4 jaw (thanks to this forum) and can get it well under .001. That is when I turned some junkyard 1 3/4 on centers and saw a taper of .002 on the OD. The tailstock had moved from it's factory setting. This is where I'm at now and want to make damn sure I know what's really happening before I move that tailstock. At the same time I want to make sure I end up with a test bar that I can depend on to align my tailstock in the future. All indicating was done with a .001 x 1.000 inch DTI.

You should check things out with a test bar before moving the tailstock (if a taper is present) you need a good stout bar with a couple of raised collars, chuck the test bar in the 3 or 4 jaw, true up the collars, then move to the tailstock end, take a .001 cut on that collar (don't touch the cross slide) wind the saddle up and take a cut on the collar by the chuck, try the .001 cut on both collars a couple of time and check the measurement of each, any variance in size will show if you have a twist in the bedways, with the exception of the truing cuts keep the cuts very small (i.e. .001 to .002) just to make sure your not getting a false reading from deflection of the work piece.

 

[epanzella]

You should check things out with a test bar before moving the tailstock (if a taper is present) you need a good stout bar with a couple of raised collars, chuck the test bar in the 3 or 4 jaw, true up the collars, then move to the tailstock end, take a .001 cut on that collar (don't touch the cross slide) wind the saddle up and take a cut on the collar by the chuck, try the .001 cut on both collars a couple of time and check the measurement of each, any variance in size will show if you have a twist in the bedways, with the exception of the truing cuts keep the cuts very small (i.e. .001 to .002) just to make sure your not getting a false reading from deflection of the work piece.

I think I get it now. To check the head alignment I have to actually turn the piece to take out all errors and insure concentricity. A number of people have tried explaining this to me but my tunnel vision prevented me from getting it. The spindle, sleeve, collet and drill rod imperfections add up and the only way to get rid of them is to turn the part. DUH! I thought I put headstock alignment to bed but now I realize I still might not be done.

 

[Don B]

I think I get it now. To check the head alignment I have to actually turn the piece to take out all errors and insure concentricity. A number of people have tried explaining this to me but my tunnel vision prevented me from getting it. The spindle, sleeve, collet and drill rod imperfections add up and the only way to get rid of them is to turn the part. DUH! I thought I put headstock alignment to bed but now I realize I still might not be done.

Yes I think you have it now, it's easy to confuse chuck/collet runout with alignment issues, and though you can force bedway missalignment out with the adjustment of the tailstock that won't do you much good when your trying to bore/turn 3 inchs and have a taper, in order to turn your work pieces taper free with or without using the tail stock proper bed alignment and tailstock alignment need to be dealt with, but the bed alignment need to be first, then move on to the tail stock.)