Rollie's Dad's method

Jim, I tried that holy water business and the machine just rusted. Magic potions are out too. Ever seen a witch? No, what works best is Maiden Juice and epithets. I learned this from the first machine shop I worked in. Almost all the work we did was close tolerance. The boss always said to put Maiden Juice on it. And then he'd cuss until things worked out. Unless his wife was around. ERS

Yessir, it sure does. There's a lot of talk about dead level, dead true, dead flat, dead straight, and plain ol' dead nuts. It's a convenient way to talk, and fine as long as nobody forgets that it means "the best I can measure with my tools." But slang has a way of taking over, so that people think it actually means what it seems to mean.

Yepyepyep. It was a bit of a stunt to get things "close enough" for the last few projects I completed. Some of them needed careful work, and I was bumping up against the limitations of my shop (and my skill). But close enough was just close enough, and the results made me proud.

Upcoming tasks are more demanding: before I'm done, I'll have had to buy a much more serious mill and probably a bigger lathe, too. $$$$$$$$ sigh. But for now, I can at least get a start on the smaller parts. Looks like I'll be able to make my "close enough" a lot closer than it was in the old shop.

Oh, hey, I'm all for it. Sacrificing a chicken couldn't hurt, either.

Pete

The idea is you provide the miniumum amount of constraint to define the geometry you desire.

My south bend lathe has the bed constrained at each end, by two bolts. So it is possible to impart twist to the bed if one tightens down the two headstock bolts, which then define the beds front/back tilt, and then tighten down the two tailstock bolts, which might want to pull the bed down to a *different* front/back tilt. Likewise when the cast iron bed expands more than the steel cabinet, the bed is going to bow up a bit in the middle.

Hardinge gets around this by mounting their bed at the headstock by pulling it down onto two spherical mounting points with heavy springs. So now the bed has been constrained to have a certain front/back tilt. It's still free to tilt left and right though, pivoting around the other axis of the two spherical mounting points.

The tailstock end has a single ball that protrudes from the botton (actually a steel bearing ball, sitting in a countersunk recess) which fits into an upwards-facing V-groove that is the third mouning point on the cabinet. This means the tailstock end of the bed now is restrained to give the bed the desired left-right tilt, and further that any differential contraction or expansion of the bed with respect to the cabinet (one is steel, the other cast iron) is permitted because of the transverse V-groove on the lower mounting point.

Of course the tailstock end of the bed is held down with the same stiff spring setup as the headstock end.

Because of the single-point mount at the tailstock, there is no way on earth there can be any twist imparted to the bed by forces that arise between the cabinet and the floor. Likewise the differential thermal expansion between bed and cabinet can not bow the bed up and destroy the accuracy.

Hardinge finished their beds mounted on those same three spots, so when they are installed in the cabinets, the surfaces of the ways agree again with the factory versions. They're not twisted or bowed.

Jim

[ ... ]

Yes -- but you would be amazed at how much more sensitive the No. 199 "Master Precision" level is. Starrett certifies it to be 0.0005" in 12" per division. They don't even make any statements about approximate accuracy for the No. 98. (I have examples of both -- not bought at new prices in either case, and I know which I would use (and did use) for initial leveling of the lathe bed.)

I put a 1-2-3 block on each of the flat ways, and put the level across those, since the Clausing has one V and one flat for the front way and one of each for the back way. (I never checked whether the narrow flats on top of the Vs was also parallel to that surface.

With the No. 199 level, you can see a change when you rest your hand on the headstock of my 12x24" Clausing, and I understand that even a Hardinge shows a shift. It is amazing how flexible a lathe bed can be.

Yep!

They simply are now properly seasoned. I believe that is a requirement for any kit castings -- that they must season under the workbench for some years. :-)

Enjoy, DoN.

I thought I said that.

I thought I said that, too.

I've

I didn't say that, nor do I subscribe to it. I also don't subscribe to ignoring level on a machine that is supposed to be leveled, making it cut straight by tweaking the bed. Unless, that is, the bed has an inherent twist that can't be removed by conventional methods. By the time machinery is in the hands of folks like us, as a community, it is often well worn. Tweaking the bed to get a lathe to cut straight near the headstock is hardly a solution when it involves twisting the bed to an out of level condition. The day comes when one will machine past the worn portion and the problems are then magnified in the opposite direction. Way too much emphasis is being placed on a lathe's ability to turn straight near the chuck, and not nearly enough on the fact that the machine should be able to work end to end with a minimum of error.

One of the negatives of pushing the bed out of level is that the tailstock is likely to be on center at only one position, so, once again, the problems are magnified. There's no way to eliminate the wear so it's not a problem aside from a rebuild.

An indicator in this case would be used for the wrong purpose. From what are you indicating? Your one and only reference point, a level bed, has been ignored, by all accounts. All conditions being correct, leveling a lathe bed properly automatically provides a proper setup. There is no need for anything else. Not until you start trying to overcome wear.

That's exactly the point. The base has much greater mass than does the bed, and it is pulling the bed out of level on one corner. There is no doubt in my mind that the minor twist in the bed could be overcome by shimming the low spot and pulling the then higher spot down to the base casting. The mass of the base is clearly greater and stronger than the bed, which is very capable of twisting, which is exactly the condition of which I speak. The base, being more rigid, will, indeed pull the bed flat. The chief reason I've never attempted the project is that I do not have access to one of the master levels, and I do almost exclusively short work. The lathe performs to an acceptable level as it sits. I don't expect my Graziano to provide the level of precision that one would expect from a grinder.

And that impairs accuracy. But even with a

Not really. Unless you choose to dismantle the bed from the base and start shimming, you gain nothing. Leveling does *not* make any changes in the bed. As I said, my lathe is high on one corner, and nothing short of shimming between the bed and base will make a difference. I can set it up with the fourth leg floating and it makes no difference. Been there, done that. I keep it adjusted so it's in contact to prevent any rocking motion of the machine, but it carries no weight. Once set, it doesn't change. My floor is 6-1/4" thick concrete.

Assuming you don't care if things roll around when placed on flat surfaces, that is. Personally, I do care. Level is not required for the machine to operate properly (same as the Graziano), but it is a good idea for other reasons, which is and has been my point right along.

And yet you continue to ignore the fact that the use of a precision level is supposed to return the machine to what most likely was factory original? Why do you suppose they suggest you start out with a level machine? I'm inclined to think it's because they know the vast majority of folks don't have access to a precision level, but they do have access to a common one. The common level gets you in the ball park, the tweaking gets you closer. That's good on a new machine, but, again, hardly a solution on a worn one. I'll stick with levels, thanks.

Harold

Betcha I wouldn't. Those Starrett Masters are just gorgeous and I want one. Most Starrett stuff is pretty, of course, and my level's no exception, but yours is the level they recommend for machine setup. The 199 is good for half a thou per foot, the 98 is graduated for five thou per foot. Oh, well!

Yeah, I always used to worry that the dog was going to sneeze in the middle of a nice finish cut and throw the whole thing off. Then I'd have to shoot him.

(Okay, yeah, the floor was the issue, not the lathe; but I did finally get an appreciation for how bendy iron can be. It took me a while.)

Yeah, except for the ones too big to fit under the bench. I get tired of barking my shins on them.

But I finally started cutting iron the other night -- made a mess, had some fun. Whee! It's been way too long.

Pete

Harold, I think we're on the same page, honestly. There's not substantive disagreement on any of this.

Yep. Folks who purchase used lathes (this encompasses a large fraction of the HSM crowd) are well aware their machines are worn when they buy them. They also are aware that they can improve performance by doing tricks that one would never consider on a new unworn machine - like shimming the bed so it cuts true near the chuck, where most of the work is done. Nobody is telling themselves "this is an unworn machine, it cuts as true as it did when it was new." The statement most often heard is, "I can make it accurate enough to get my task done, in tolerance."

And that is probably the *most* important issue - a base that does not move over time. I would be very curious to see how the graziano cuts for cylinder, using the two collar test.

Do they use spirit levels when manufacturing machines? I know that the Connolly book discusses their use in some applications. The Moore book has literally no mention of them.

Jim

Stress is force and strain is displacement, right?

If there were no gravity, one would merely puddle mortar or expoxy at each foot, allow to cure, turn gravity back on, then begin operations. This would assure the lowest stress mounting, and you'd figure that would produce the lowest strain mounting.

Without a gravity switch, would it be right to have adjusters or shims at each foot such that a separate piece of shim stock, the same size at each foot, could be pulled with a tension meter, assuming friction conditions were the same at each foot, to even the load on each foot, or conform these loads to a chart of the way they should be, as a first step?

I get this idea from taking off the front of our dryer. It's a big floppy assembly with the panel off, but when in place, it's 250 pounds and more, solid. So I was thinking that when I was replacing the rollers for the dryer, the thing to do would be to adjust the feet so the rollers contacted the drum evenly. The dryer's four feet, two fixed rear and two adjustable front, are redundant.

It's a question of kinematic redundancy. A lathe with two headstock feet and a single tailstock foot would be ideal. It seems you could plunk such on a level floot and go right to work because three point contact defines the position.

--Doug

In my mind, the purpose of a level is to duplicate the condition in which a machine tool was manufactured. That does not imply that they use levels in the process, but allows one to duplicate a condition after the fact. If you level a bed, it will be in the same plane as when manufactured, at least in theory. I think both of us would agree that it could lean in any direction (but remain in a flat plane, regardless of angle) and still perform as intended. The purpose in leveling is to establish a datum plane, or point, so one can make relative measurements (or comparisons) with ease. Machines that are not level make it very difficult because of the constant compensation for the degree of tilt. Beyond that, I don't see it as a necessity. In fact, I recall advising one fellow that he could set up his small lathe at an angle such that he could access it easily from his wheel chair. So long as the bed is not twisted, it makes no difference that it leans one way or another. What say you?

Harold

A bit different way of looking at things for setting up lathes in good condition.

Think about the way many bench lathes are made. The bed is machined when it is mounted on a flat level surface but BEFORE any extra bits are bolted or mounted on it. In this state the bed is pretty symmetrical so there is little reason for twist to develop if the flat surface on which it is mounted is tilted.

A plethora of strange shaped bits is then bolted or mounted on this nice true bed. It is these asymmetric bits and the drive belt tension that produces small asymmetric forces that can contribute to bed twist. Note that this is true even when the lathe is dead level so that to return the bed to its parent unstressed shape a very small counter twist needs to be applied.

If the lathe mounting surface is flat but not level there will be a small change in these asymmetric forces and this would require corresponding correction to the counter twist setting (in some cases the necessary correction forces would reduce). There is no reason why this correction should be any less complete than in the dead level case.

This supports the argument that a carpenters level is more than adequate for basic levelling.

If you're fortunate enough to possess a precision level a useful measurement is to simply locate it along the the crossslide and note the bubble movement as the carriage is traversed from end to end. It doesn't matter whether the bubble is centred or not - the ONLY thing that counts is how much it moves.

This is an interesting measurement as it makes it possible to directly measure the effect of asymmetric weight distribution on bed twist. I'm lucky to have a slightly home brewed electronic level which can read down to +/-1 min arc full scale. On a Boxford ME10 (pretty similar to a 9" swing Southbend) a 14lb weight attached to a

12" arm overhung from the tailstock end of the bed - i.e. 168 lb/inches twists the bed by just over 2 min arc in 12" of carriage movement. 160 odd lb/inches is a severely asymmetric weight distribution Because the bed twist axis is a about 6" below the headstock axis, 2 min arc corresponds to a toolpoint shift of about 0.0036" which is a serious amount. However even this much can be cancelled by appropriate adjustment of the tail stock end mounts.

The interesting thing is that the correct setting of this adjustment is almost unaffected by reasonable deviations from level - a 1 deg error only changes the asymmetric force by cos 1 deg - less than 2%.

Incidentally I have based these comments on forces rather than shim thickness or movement because it is the balance of forces that determines the bed state - the actual change produced by a 1 thou shim depends on both the relative stiffness of the bed and its mounting surface.

It's not a recommendation to rely on a precision level for final setting up - I still think the best method is a tenths clock bearing on a reasonably straight and round test bar with eccentricity averaged out. Jim

Geometrically, you are entirely correct. I do find, however, that machines running coolant prefer to be level.

A small point, not germane to all, but helpful if you are running coolant..

Another data point is: if I touch off on the spindle nose with a good indicator mounted in the toolpost, measuring horizontal deflection, it takes only one finger pushing on the headstock casting to get the dial swinging by several tenths. This is on a 10L SB machine.

Jim

Yep! Yet another *good* reason to level machines. One should also consider that many machines have oil reservoirs that should be kept level.

Harold

That's the point: shim it so it's level and not twisted, rather than shimming it to deliberately twist it to correct another misalignment.

In this case often the shimming is done to counteract the effects of wear, so the machine turns and bores straight.

To put this in perspective, the shims involved in getting a small floor mounted lathe dialed in for the last thou of accuracy over, say, a six inch long part, might be

0.015 inch at the legs. So the shift in the bed is going to be a couple of thousanths or so. Not much.

Jim

Pete, I've had to rely upon Rollie's Dad's method and it has worked very well for me.

Like others have said, level the lathe using a precision level. Then, if you feel a need to check headstock alignment, use R-D-M.

In my case, I bought a used lathe that was in good shape, but I knew something was amiss when I spotted a crude shim between the headstock and the bed. Many days of bluing and scraping along with R-D-M to check my work eventually paid off. I really enjoy using that lathe, now.

Regards,

Orrin

I think that is true. I'm just trying to imagine how for example the south bend beds were made.

First cast, then probably the ways were roughed out on a planer.

At this point there was possibly a level used on the planer bed - but my guess is they first got the mounting surfaces all in the same plane, then flipped it over and did the ways. So when you were done, if the planer bed were level then the part done on the planer would be also.

Then the bed goes to the scraping department - and there I don't know if they simply used spotting gages or if they make more sophisticated tests. I know in the moore book they describe the use of full length test gages and even more intricate tests like autocolimators and so on. The implication is that maybe the level things before they start, but that's only the beginning, a spirit level even like the starrett master precison on simply won't come close to giving the accuracy for a machine like that.

Reading between the lines, I get the impression that the south bends were pretty much Chevys, and something like a moore would be a deusenberg.

And that the things the moore folks worried about were even conceived by the guys at the south bend plant.

Jim

[ ... ]

Based on what I have seen in the South Bend _How to Run a Lathe_ book, the ways were cut -- all in a single pass (except perhaps for the under surface used by the carriage gibs and the carriage lock), using an amazing stack of cutters on a single arbor (with lots of bearings holding it right), so all the planes of the lathe bed were cut at once, and in a constant relationship to each other.

O.K. It is not in the 50th edition (which is the one which I can quickly lay my hands on), but is either in a later edition (light gray cover, instead of the black on the 50th edition), or perhaps in the similar book from Atlas. But I remember the cutters almost vanishing under the flood coolant.

And another shot of the roughed lathe beds being stacked for seasoning prior to the fine cuts and the scraping.

Now -- I suspect that much larger machines *were* made on planers, but for those of the proper size, the horizontal spindle milling machine would do a lot more a lot quicker. In particular, the number of setups to get the right angles on the bed ways (for the prismatic ways) would make a planer take a lot more intense operator interaction. And the time to cut to a near-finished state would be a lot longer than that horizontal milling machine with its cutters awash in chips and coolant.

Of course, later machines, such as my 12x24" Clausing (1957 vintage) were flame hardened and then surface ground to final dimensions. (They probably were still run through the horizontal mill before that.)

I seem to remember in one of the reconditioning books a "level" made of a trough of water the length of the bed. Harder to read with a

1 to 2 foot bed length, but when you get one of those monsters with chairs on the carriage for the operator, it should be quite adequate.

Yep. The same (even more so) for the Atlas ones. It is hard to believe that Clausing and Atlas are the same company, comparing my lathe to the old Atlas/Craftsman 6x18 which I also have. :-)

You mean "were *not* even conceived", perhaps? That, I can agree with.

Enjoy, DoN.

I would think that would be the case, perhaps with some seasoning time along the way after roughing. I also wouldn't discount them having been machined on large mills at a later date. Shapers and planers fell out of favor due to inefficiency. Perhaps no seasoning, but heat treatment ( stress relief) instead. One thing for sure, the machinery on which they were made was all leveled to a fine degree. Not that it makes any difference to how a part would be handled, but to insure that the precision surfaces of the manufacturing machinery were in a common plane, yielding straight cuts. To ease handling, I can't imagine that the beds weren't parallel with the mounting feet, which is the only thing that really matters. So long as the two surfaces are parallel, leveling the top portion would, in theory, make a proper setup. Considering you own a SB, and I don't, nor have I even seen one in years, perhaps you could tell me if the ways and base are, indeed, parallel.

Yeah, like that, only level isn't the consideration so much as

*parallelism*.

Maybe for the SB, but not for modern machines. I dare say there is *no* scraping involved, especially when flame hardened ways are involved. Grinding long ago took over that level of precision.

While we're discussing lathes, it's more than obvious to me that Bridgeport mills had *no* scraping in their making. Flaking, perhaps, but they were either milled or ground to size, with flaking for oil control. You can see that clearly when you look at one that has not worn excessively.

and there I don't

Agreed.

If even that. I am not a fan of SB lathes, never have been. I'm turned off by their insistence of sticking to old technology, flat belt drive, sleeve bearings, clutches that needed to be engaged by turning knobs instead of rapid levers, like that. They may have been a wonderful machine at some point in time, but failed to keep up with technology that made machines better, and easier to operate. No offense intended. It's just a personal bias. I realize that they are worshipped by many.

Incidentally, your mention of a Chev as compared to a Duesenberg. Wow! You really know how to pull at my heart strings. While likely a pig to drive, the J model Duesenberg is one of the cars that has always haunted me. At one point in my life I was involved in classic cars (no, not '67 Fords, which will *never* be classics, regardless of what you may read), but the real thing------true classics. I owned a '36 convertible coupe Auburn, supercharged, with a Columbia 2 speed rear end, plus a '37 Cord, Custom Beverly Sedan, supercharged, which was a one of a kind car, authenticated by the Cord historian of the ACD Club. The car had a Berline interior, and was known to be the sole Cord so turned out by the factory. At any rate, I always coveted the Duesenbergs, which, at the time, weren't all that expensive to acquire, very unlike today. I had the privilege of working on one, a '29 Murphy sedan. It's hard to imagine the beauty of the engine, a straight 8 with dual overhead cams, 420 CI displacement.

I've mentioned this before, but it's worth repeating. In the late 50's, Sperry Utah purchased a new Fosdick jig borer. It was in addition to the Moore they already had. You've never seen such scraping in your life. Having watched the machine being installed, it was a perfect opportunity to see scraping at its best.

Moore, from all indications, more or less set the standard for high precision. Their lead screws were represented to be precise over 12" to something in the low millionths. Lovely machines, they were.

Harold