Lathe Leveling

[Stan]

I changed lathes a couple of months (Logan 1825) ago and spent a couple of weeks getting the lathe bed level. I don't have a 'precision machinist's level' but I do have a Starrett 98 (.005"/ft).

Every day or two I would rearrange shims under the four feet trying to get the bed level across the bed. I wasn't worried about lengthwise other than approximate. When it sat still for a few days I replaced the stack of shims under each corner with a stack of not more than two shims and rechecked in a couple of days.

On Tuesday, this week, I decided to take a long test cut and got .006" taper. On checking the level, I found the front right corner low 1 division on the level (about .010" on the width of the lathe) A .010" shim under that corner reduced the taper to .003" and by yesterday it was down to .002" but the headstock end was now out half a division.

Lathe manufacturers emphasize the importance of leveling the bed and I have my lathe sitting on a concrete floor and I have trouble getting it level or to stay level. What do people do with lathes on wooden benches or wooden floors? It is apparent that a small amount of twist can cause big problems.

Maybe my method is wrong. I start with a piece of 2" 6061 aluminum round bar stuck in a 3 jaw chuck. I take a toughing cut to get it round and then take cuts of .005" or .0025" and let it cool before measuring. My reasoning is that 2" with small cut should have negligible spring and I should be able to have 8" or 10" out of the chuck. It is quite easy to change the taper from too big at the chuck end to too small at the chuck end by changing the level of the tail stock end.

Anyone with a good suggestion?

 

[Bogstandard]

Hi Stan,

I had a lathe very similar to yours, based around the same design.

http://www.homemodelenginemachinist.com/index.php?topic=137.0

The last pic in the set.

If yours is on an original manufacturers pressed steel stand, unless it is of the internally reinforced type, they are renowned for twisting, which will then be passed onto the lathe.
The way around this is to put the lathe on a very rigid base, ie, a good thick wooden top or a purpose made stand, as mine was, made from 3" square section thick walled tube, with a lot of cross bracing. So coupled with the rigidity of the frame and the lathe bolted to it, nothing is going to twist.
I shimmed mine between the original lathe feet and the riser blocks that I made for it. It just never moves.

The way you are checking does show that the twist in a lathe does affect the way it cuts. I only have about 4" or 5" sticking out, over that sort of figure, gyroscopic and centrifugal forces start to move the bar out of position in relation to the tool, depending on chuck speed. I would try as a finishing cut of 0.001" to 0.002" as even a little larger cut can deflect the bar.

Even resting your hand on the lathe when taking a cut can cause a taper.

Everything depends on rigidity.

John

 

[rake60]

A lathe is a finicky animal.
Bed twist is it's worst enemy when it come to cutting straight.

It can be set up perfectly and 3 days later it tapers to the chuck.
3 days after that it might taper away from the chuck.

Being a turning machine any error is doubled.
If there's .002" twist in the bed the measured cut tapers .004"

Different metals expand and contract at different rates as
ambient temperatures fluctuate. The lathe bed and frame is usually
cast iron, but the hardware holding it together is normally steel.

There's no such thing as a perfectly set up machine.
Of course we want to get it as close as possible, but it will
change with temperature, weather and sometimes I think they
just have MOODS!

Rick

 

[Stan]

An update on my lathe level. When I came home this afternoon, I checked the level and it was perfect at both ends. I put a fresh piece of 1 1/2" 6061 in the chuck with a 6' stick out. I took a couple of cuts and then put in a sharp HSS bit and took a .002 cut and then measured taper. It is now about two tenths ( .001 mic feels a little tighter when moved from end to end).

Now it is like prostrate cancer - wait and see if surgery is required.

John: this cabinet is pretty substantial. There is a difference in shipping weight of 355 pounds between a bare lathe and the cabinet model. However I am sure it is not as solid as the cast iron legs under my previous lathe (Logan 820). If the end closest to the tail stock was consistently larger I would expect spring which is so common on small diameters. When it changes from end to end with changing shims, I am convinced it is bed twist.

Rick: I have a basement shop and in some ways it is a pain but with central heat and A/C there is very little change in temperature to affect my machines. In fact my old lathe stayed rock solid for six years in that position.

As a matter of interest: About a year ago I bent the spindle in my 820. After giving up on repairing it, I bought the 1825 and planned to part out the old lathe. About two weeks after buying the new one, a headstock showed up on ebay with a complete spindle assembly in it. I bought it and replaced the complete assembly and the old lathe was back on spec. I put an ad on Craiglist and sold it, thus keeping another old piece of iron alive.

 

[Bogstandard]

Stan,

A man after my own heart.

These old lathes, if looked after, can produce some astonishing work, and can go on for years.

I honestly think it was all to do with the weathering of the castings, where they were left outside for a few years to stabilise, then they were machined. Nowadays, it is cast one day, machined the next and shipped the day after. Just like all this kiln dried wood, compared to the old outside seasoning. No comparison.

John

 

[Foozer]

Getting the lathe level, or as the terminology finally set in, as I read it, bed twist takes priority over the term level.

And so after plowing down some blackberry bushes and getting the truck stuck in the mud figured getting the lathe bench "level" would be a good way to spend some time. Now I'm a cheap'o and putting out money on a precision level that will get used little is 1. too easy and 2. no real fun at all. So on to the Red Green approach.

Bench is wooden, 4x4, 2x12 and double 3/4 ply. Legs have angle iron attached which bolt via studs in the concrete floor. using a regular carpenters level shimed the legs to achieve a level top within the capacity of the bubble and my vision. Bench itself is sturdy enough that all I really did was remove any movement caused by the uneven floor (which being an old barn it is way out of whack) But is it level?

Took an old laser pen, gutted it and mounted it within a piece of stock, this mounted to a V-Block and then rested upon a known piece of good flat stock.

From one end of the bench shined the laser dot onto a target some 18 feet (5.5 meters) marked it, (bottom of duct tape, I did say Red Green) repeated the process from the other end of the bench with some 2 inch (5cm) difference.

So if the process is sound, seems like it, wont tell me if the bench top is "level" but it seems that any twist in the bench top should be shown by the differance in laser dot shinning to the same target from either end of the bench. Sure there is some math that would answer what the amount of twist is based on the data but that just makes my head hurt. Tomorrow will add/remove shims as necessary to get the dots together.

 

[joeby]

The subject of machine leveling has been beaten to death over the years, but the fact of the matter is that the machine tool doesn't care all that much if it's level or not. You could bolt your lathe to a wall and it would do just fine, the important thing is that the ways remain parallel to the spindle center line. This is where leveling becomes important, the level is your reference to determine the misalignment of the ways. This is the reason that "Rollie's Dads' method" works.

Interesting that the EDM machine that I used at my last job sat on rubber pads which in turn were placed on top of the base/dielectric tank. The leveling requirements were to make the machine set on all four feet solidly and be sure the dielectric would drain from the work tank. That machine was capable of holding .0002" repeatably if the other concerns such as room temperature, electrode tolerance and such were taken care of.

Kevin

 

[Foozer]

The subject of machine leveling has been beaten to death over the years, but the fact of the matter is that the machine tool doesn't care all that much if it's level or not. You could bolt your lathe to a wall and it would do just fine, the important thing is that the ways remain parallel to the spindle center line. This is where leveling becomes important, the level is your reference to determine the misalignment of the ways. This is the reason that "Rollie's Dads' method" works.

The amount of topics I have read proves your point

the important thing is that the ways remain parallel to the spindle center line.

In the effort to obtain that requirement I just start from the ground up reducing the margin of error as I go. This machine is small, an old Craftsman 109, so it gets bolted down to the bench. The RDM process is of course simplicity, elegant, yet however the old brain that came up with it surely went through and dismissed those steps as unnecessary that I am embarking upon.

So I can now say that the bench top may not be level, but it sure taint twisted.

Its all good fun

 

[tel]

Yep 'leveling' is a misnomer, what we are trying to do is get any twist out of the bed. Now, as far as I am aware, the human eye is still the most precise instrument there is, so what you do is drop the tailstock off - lay two parallels across the bed at the extremities, and eyeball across 'em. That's how I set up the Myford in its current position some 20 years ago, and it still turns true.

The old time carpenters used to use this system for getting the twist out of bits of furniture, with a pair of 'winding sticks'

 

[joeby]

Foozer,

As could be said about this hobby, it's all just means to an end, the end result is what really counts. RDM surely was brought about by the tools that were at hand, the principle behind it remains the same. Possibly, the only advantage RDM has over your method is that it also would indicate a bow or sag in the ways, which shouldn't exist anyway, being that a lathe bed is built to resist that type of problem.

I also have a 109 that I got "horse trading" probably fifteen years ago; but never really did much with it. The plan was to make it portable, to an extent, by bolting it to a steel plate that had been Blanchard ground and using three legs under the plate. The three point support combined with the heavy (about 1 1/2" thick) plate should have prevented twisting and enabled the lathe to be used on a bench and either moved to the side or lifted off and stored elsewhere when not in use. One of these days....

Kevin

 

[Andy_B]

Tel

My dad taught me about winding sticks when I was a wee lad tagging behind him in his cabinet shop. Have been using it on furniture for 50 plus years but never transferred the idea to resolving lathe twist. : Thanks for teaching an old dog a new trick. ;D

Andy

 

[tel]

It works fine mate, and the longer the 'sticks' the better the result

 

[steamer]

Rollies Dad's method


http://www.neme-s.org/Rollie's_Dad's_Method.pdf

It will take a minute or two to wrap your head around the concept, but it needs very little in the way or equipment. An old auto strut rod, as long as it is ROUND, will work fine....doesn't even need to be straight.....and a good dial indicator, and a and a magic marker

Check it out and think about it a bit...it does work really well.

Dave

 

[Foozer]

Yep 'leveling' is a misnomer, what we are trying to do is get any twist out of the bed. Now, as far as I am aware, the human eye is still the most precise instrument there is, so what you do is drop the tailstock off - lay two parallels across the bed at the extremities, and eyeball across 'em. That's how I set up the Myford in its current position some 20 years ago, and it still turns true.

I'm an extremely abstract thinker so my words chosen often do not clearly relay the intent. One gets used to it after this long on the planet

Misnomer, thats the word. The eye is indeed a precision device, mine however are somewhat skewed, keratakonis, for the past 40 some years. What I think I see is not always what it really is at the smaller resolutions. Just a matter of finding methods that accurately scale up items to be viewed.

Foozer,

As could be said about this hobby, it's all just means to an end, the end result is what really counts. RDM surely was brought about by the tools that were at hand, the principle behind it remains the same. Possibly, the only advantage RDM has over your method is that it also would indicate a bow or sag in the ways, which shouldn't exist anyway, being that a lathe bed is built to resist that type of problem.

A case to which "The end justifies the means" is valid

Rollies Dad's method


http://www.neme-s.org/Rollie's_Dad's_Method.pdf

It will take a minute or two to wrap your head around the concept, but it needs very little in the way or equipment. An old auto strut rod, as long as it is ROUND, will work fine....doesn't even need to be straight.....and a good dial indicator, and a and a magic marker

Wasnt bad, actually liked the abstract nature of the concept. The method counts on having separate "cam lobes" rotating in different points of space the difference between the 2 is then simplicity to visualize. Tried it last night after setting the bench top to remove any detectable twist. Well at least the laser dots as indicated from each end of the bench converged on the 1/4 inch dia target 18 feet away.

RDM gave results of 0.003 off in the vertical and .0045 in the horizontal. A little putt'ng with it today may reduce those numbers by half. Not the nature of the little beast to hold perfect nor do I expect it to do so.

Thanks to all for the feedback