Hardinge tailstock misalignment causing taper.

Hello!

I am currently going through the machine work program at the local community college. Most of the manual machines we have are pretty hammered. One of the few decent ones left is a Hardinge HLV-H. It's virtually the only lathe I bother using anymore (as both 10EEs are out of commission).
Anyway, today, I was trying to make a couple parts that only had one or two thousandths for tolerance over a length. As the title of this thread explains, I was getting some taper in the cuts. My instructor and I quickly came to the conclusion that it was from the tailstock being misaligned. We quickly disassembled it to find shims between the tailstock and the gib that slides against the bed's dovetail. We put the gib on the surface grinder and smoothed it out (non-sliding side) as to try and remove any high spots. After re-shimming and indicating inside the tailstock quill, it was still a little off. However, tightening the tailstock bed lock was showing us more variation than we were getting while doing the measurments. It made me wonder about the overall accuracy of the system. If tightening the tailstock lock could shift the whole thing over a thousandth or two (or more), how can I be sure everything is lined up when I'm trying to hold tight tolerances? Is it just an inherent part of the design? Does anyone have any advice or experiences with this? To me, it seems like downforce applied to V-ways would be more accurate than a push from one side of a dovetail.

Consider the possibility that the front and back edges of the bed are worn, especially in the area most used, near the headstock end. I am guessing you cranked the carriage as far to the left as it will go, so that you could clamp the tailstock close to the headstock to indicate the Morse taper. Then the tailstock might have been on a somewhat tapered or curved part of the bed, and not lining up the same as when it is farther from the headstock.

The variation when tightening the tailstock clamp might be from a non-straight part of the front edge of the bed, or wavy or dirty shims that compress like springs under variable presure. If the gib and/or the bed is not straight, the gib might be bending.

Then there is the question of how tight the tailstock spindle is in the tailstock bore, and how parallel it is with the bed. Different extensions of the tailstock spindle might change your alignment.

Larry

Did you check for level across the ways under the chuck, then behind the tailstock? See Chapter II, pps. 15-16 in "How to Run A Lathe" by South Bend Lathe Works. It sounds like the lathe might be out of level, or racked across the ways.

Hardinge HLV-H lathes don't work that way. The entire bed rests on 3 spring-loaded thru bolts with a ball and socket arrangement. You just adjust the feet to take any rock out of the cabinet and let it go. There may be other reasons for leveling (such as keeping your tools from rolling around), but accuracy isn't dependent on how level the Hardinge is.

You beat me to it, Jim.

I was gonna say, you could twist that lathe
cabinet up like a pretzle and it won't change
what the bed looks like one bit.

Now on to the real question, why is the part
being turned taper?

First off, what is the setup? Are you truly
turning the part between centers?

Or you you holding one end in a collet or
chuck, and putting the other end in the center
in the tailstock?

Jim

Thank you for all the input. I just had one end held in a collet and the other on the live center. I got the taper in a few different parts (of varying length and diameter). Would the shimming of the tailstock gib be the most likely culprit? Where else should I be looking to try and chase down the problem? For measuring, my instructor just had me put a dovetail holder in a collet with the indicator attached at the back end. We then put the point inside the tailstock quill and swept it around. There was quite a bit of movement vertically, too. However, from one side to the other (what we assumed was causing taper), we were getting a .001-.002 difference.
Anyway, the shims behind the gib seemed like the most logical way to dial it in, but the material used for shims was garbage. Any more suggestions? Thanks again for all the help.

I have been wanting to purchase a HLVH. I am interested in this subject also.

Based on your latest comments regarding varying length and degree of taper, I would say that you have a lack of rigidity either in the material, or the tool holding setup. It could also be that you are using carbide and trying to take too shallow a cut. Once the tailstock is locked in position, if it were responsible for the diameter changes, it would produce a constant taper. After all, that is one way to deliberately turn a taper. The fact that you are seeing changes in the amount and location of diameter deviations, suggest that the tool is alternately cutting and deflecting, or that something is moving. A radius carbide cutting insert requires fairly high pressure against the part to actually cut. Try using a HSS cutting tool with a small radius, or a carbide insert with the smallest radius you have. Increase the diameter and stiffness of the material you are turning (say 1" leaded steel) and check your compound for rigidity.

You can tell what is going on with the machine
by doing two things:

1) chuck up a two inch diameter piece of aluminum
and turn the OD down a few thousanths with a
sharp HSS tool, *without* touching the
cross slide during the cut, and *without* using
the tailstock AT ALL.

2) turn a a similar part truly between centers,
driving with a dog. NO collet, NO chuck.

Report back with the deviation you see from
cylindricity for the two different versions.

Hint: Holding a part in a collet *and* with
a tailstock center is going to give some
pretty muddled results. It's not usually
done.

Jim

Thank you for the pointers, Jim(s). As I mentioned, this is the lathe at school. So, while it isn't used in production, it does have the occasional lunkhead at the dials.

The pieces I was turning (.375 and .250 1018 CRS) showed a consistent taper of about .002 over 1.5". Both times, I did have the stock in a collet AND on the tailstock. Until now, I wasn't aware that the two usually aren't used in conjunction. My instructors have never made mention of it. Is there a particular reason why they shoudn't be used together?

As for chasing down the problem... I was using HSS with a fairly small radius. Speeds, feeds, and surface finish all seemed in order. I'll try the chuck with some 2" aluminum as suggested. I don't think the machine at school has a plate for driving dogs, but I'll check. However, I still think the problem is coming from the cheezy shims that were in the tailstock. I'll keep looking! Thanks again for all the help. Part of the reason I'm eager to figure it out is because I have a Hardinge TFB in my shop that I've yet to run. I'd like to learn all the nuances so I can get the most out of it.

When I went through this alignment on my tailstock
it was apparent that the 4 allen bolts on the front
part of the tailstock allows one to take out any
horizontal mis-alignment. I don't agree that the
Hardinge tailstock is non-adjusting.

The hardinge tailstock is non-adjustable
in the *vertical* direction.

Did you do the two tests I suggested above,
and what was the result of the one performed
without the tailstock please?

Jim

I am most interested in this post because I suffer a similar problem with my HLV (not -H) and it has been driving me crazey for over a year now. I have gebneraly re-furbished all the easier stuff (main spindle bearings etc. but no scraping in) and doing the 2" aluminium test I get brilliant results -- infact I actually I can use a piece of 1.5" fcms sticking out 5" and it is consistently parralel within =/- 0.0001 using a sharp small radius HSS cutter. So I am happy about the head alignment. But as for the tail -- turning between centres I found a taper of 0.005 over 6", I found odd paper shims behind the tail stock gib so I removed them now on my older HLV the 4 screws on the front side seem to only serve to hold the gib hard up against the tail casting -- because there are clearance holes in the casting and the screws are threaded into the gib. My gib has the centre section that bears on the bed relieved presumably so that it gives a 3 point clamping arrangement with the backside lock. Here is where I get the similarity with ions82 the horizontal alignment of my tailstock centre varies depending on the amount of locking force applied. I have double checked all my ram fits etc. and have reamed out the 2mt socket actualy I now taking the horizontal alignment readings off the actualy tailstock body casting. What really gets me is that I can set up the alignment by a combination of added shims behind the gib and adjusting the locking lever so that it always gets pulled up against the stop pin but if i re-test the next day its wrong again also I can read a thou or 2 horizontal movement of the nose of the main casting by leaning hard on it. I have checked the casting for hairline cracks but cant find any so I am really stumped for an explanation -- I do hope one of you can help as I said its really driving me nuts.

cheers Nigel

Hello to the comunity from greece

I am trying to align the centers of my newly restored english kl-1 (hlv-h).
Before i ask any question i ll try to explain (with my not so good english) what i ve done since now to bring my lathe centers as close to 0.

For the headstock:
First step was to align the headstock that was off when i replaced it back to bed after the restoration .
Found that the locking pin that is behind the headstosck casting not tram the headstock perfect so i tried to bring it as closer to 0 unscrewing/screwing the 4 hex bolts that holds the headstock to bed taping with a plastic hammer .
The way i was trying to tram the headstock was of taken light passes on a 30mm diameter 15cm lenght 7075 aluminum each time i was unscrewing and taping the headstock.

My best result was 0.01mm or 0.0004inch off from one side end to the other on that 15cm lenght aluminum bar.

For the tailstock:
Tailstock was the second step and since the machine was 35 years old wasnt expected to by quite true and aligned with the headstock .
I made a quick test with centers on both headstock and tailstock and the results was very disapointing.
The gib was is a bad shape , corroded and slitly bended so i decide to make a new one from cast iron ... the new custom made gib milled and grind to 11mm thick (the old one was 11.05 so i made it 11.00 so i can align it with shims)

With the new gib installed ,tailstock spindle total open and locked i made a test using mt2 test bar bringing a test indicator back and forth so i can see if the tailstock spindle is allined with bed .
The results was quite satisfied ... 0.005mm or 0.0002inch off on my 12cm lenght test bar


3rd step since i have headstock and tailstock alone as close to 0 with bed is to align both centers .
As i said it is a 35 years old machine and the bed is obviously more worn in the area near headstock .
The method i used was to plac the test bar between centers so to shim the tailstock properly till i read 0 in my test indicator .
Problem is that due to the worn bed when i am on center (with a 0.1 shim between tailstock casting and gib) and the tailstock to the right side of bed , i am about 0.04-0.05mm off when move tailstock to the left side of bed close to headstock .

Since its imposible to regrind the bed (lot of money and trouble )my first thought is to cut shims from 0.01mm to 0.05mm and use them between tailstock gib and bed dependind on the position of tailstock (thicker shim as closer tailstock is to headstock )

My questions are .....
Have you ever use a shim like this so you can adjust / align the tailstock or do you have any other suggestion on this problem with worn bed and faulty taistock allignement ?
What readings/accuracy you get from your old hlv-h ?

Thank you very much ,

Kostas.

Google 2 collar test, practical machinist ans and test it that way. Use a micrometer to measure it, not a indicator. .0004 in what distance and which end is bigger? Also you need to make dry or free passes. Meaning cut the 2 collars 6" apart with .1/2" wide collars and under cut the middle and about 1/4" in frount of the collar nearest the chuck. You cut the tailstock end of shaft and the disingage feed and manually crank the saddle foward to the 2nd collar and engage feed and cut that collar, when your tool passes thru that collor into the 2nd relief shut off feed ans stop spindle. DO NOT TOUCH CROSSFEED HAND. LEAVE IT IN THAT POSITION. Then carefully crank saddle to Tailstock end and do not feed crosslide in and tiur on spindle and repeat the procedure. This will eliminate any chace of push away from tool.

The tail stock end should be small .0001 to .0002" in 6" . Did you grind the bed and srcape saddle? If you scraped tailstock it will be lower then head and you will have to either lower the head by scraping the 4 pads where it bolt down or Moglice the tailstock to raise it up. You can You Tube Tom Lipton Oxtool as he shows these tests. Do that and tell us what happens. USE a. Digital or pressure micrometer and not the indicator.

Kostas, welcome.

Next time you post, please start a new thread instead of hijacking an old one.
JR

A thou (or 2) over a couple inches ain't bad. Rather than try to "fix" the machine, just set the compound to turn a straight piece or sand into tolerance. Expecting a school machine to be perfect and stay perfect is unreasonable.

Not the "ideal" solution, but neither is working for a living. If you're serious about making parts, choose the option that makes parts, every time, over the ideal solution. It's a choice you'll run into, a lot.

The attached photo is my solution to turning a taper on an HLV-H. Have not used it yet, but I think it will work fine when needed. I see no reason it could not be used to correct a machine that was worn or abused. The point was ground from 1/2 high speed steel blank but could be made of carbide if a lot of work was anticipated.

Thanks all of you for your suggestions

Richard King

Google 2 collar test, practical machinist ans and test it that way. Use a micrometer to measure it, not a indicator. .0004 in what distance and which end is bigger? Also you need to make dry or free passes. Meaning cut the 2 collars 6" apart with .1/2" wide collars and under cut the middle and about 1/4" in frount of the collar nearest the chuck. You cut the tailstock end of shaft and the disingage feed and manually crank the saddle foward to the 2nd collar and engage feed and cut that collar, when your tool passes thru that collor into the 2nd relief shut off feed ans stop spindle. DO NOT TOUCH CROSSFEED HAND. LEAVE IT IN THAT POSITION. Then carefully crank saddle to Tailstock end and do not feed crosslide in and tiur on spindle and repeat the procedure. This will eliminate any chace of push away from tool.

The tail stock end should be small .0001 to .0002" in 6" . Did you grind the bed and srcape saddle? If you scraped tailstock it will be lower then head and you will have to either lower the head by scraping the 4 pads where it bolt down or Moglice the tailstock to raise it up. You can You Tube Tom Lipton Oxtool as he shows these tests. Do that and tell us what happens. USE a. Digital or pressure micrometer and not the indicator.

Click to expand...

I ll do that 2 collar test .
Supose this technin eliminates the possibility of a wrong measurement , is it?

neilho

A thou (or 2) over a couple inches ain't bad. Rather than try to "fix" the machine, just set the compound to turn a straight piece or sand into tolerance. Expecting a school machine to be perfect and stay perfect is unreasonable.

Not the "ideal" solution, but neither is working for a living. If you're serious about making parts, choose the option that makes parts, every time, over the ideal solution. It's a choice you'll run into, a lot.

Click to expand...

I mostly do small parts for r/c models , some times tiny parts with 2-3mm ID and 5mm OD (bore reducers) so i cant think another way to build an accurate part than a perfect aligned tailstock for drilling and reaming operations.

The attached photo is my solution to turning a taper on an HLV-H. Have not used it yet, but I think it will work fine when needed. I see no reason it could not be used to correct a machine that was worn or abused. The point was ground from 1/2 high speed steel blank but could be made of carbide if a lot of work was anticipated.
Attachment 228503

Click to expand...

That is a good idea , but problem on drilling - reaming operations will still exist.

Kostas, welcome.

Next time you post, please start a new thread instead of hijacking an old one.
JR

Click to expand...

Sure , no problem .