PDA

View Full Version : How to Shim a Few Tenths



Paul Alciatore
11-08-2009, 04:12 AM
My question is just that. How would you shim something a few tenths? I know common aluminum foil is about seven tenths, but what If I need a finer adjustment? Some way to work in two or three tenth increments. And I am not willing to spend $50 or $100 for big rolls or shim stock. I'm looking for the bargain basement solution.

I know someone will ask exactly what I am up to so I'm trying to change the angle of my headstock by a few tens of minutes. I am not experienced at scraping and do not want to try that at this point. It's a SB9 with the headstock resting on a Vee way at the rear and a flat way at the front. Since the rotation is so small, I figure a shim on the front of the Vee at one end of the headstock and on the back of the Vee at the other end, will rotate it by a small amount. I calculated about 0.001" at each end, but I am sure some trial and error will be involved. And, yes I know I will raise it a little bit and will have to also shim the tailstock. Actually, that is part of the reason why I want to be able to work in tenths as the two may not need exactly the same amount.

Timleech
11-08-2009, 04:29 AM
My question is just that. How would you shim something a few tenths? I know common aluminum foil is about seven tenths, but what If I need a finer adjustment? Some way to work in two or three tenth increments. And I am not willing to spend $50 or $100 for big rolls or shim stock. I'm looking for the bargain basement solution.

I know someone will ask exactly what I am up to so I'm trying to change the angle of my headstock by a few tens of minutes. I am not experienced at scraping and do not want to try that at this point. It's a SB9 with the headstock resting on a Vee way at the rear and a flat way at the front. Since the rotation is so small, I figure a shim on the front of the Vee at one end of the headstock and on the back of the Vee at the other end, will rotate it by a small amount. I calculated about 0.001" at each end, but I am sure some trial and error will be involved. And, yes I know I will raise it a little bit and will have to also shim the tailstock. Actually, that is part of the reason why I want to be able to work in tenths as the two may not need exactly the same amount.

Gold leaf? It comes in a range of thicknesses, down to almost nothing. There's also what I think they call 'metal leaf' which is imitation, cheaper & thicker.

Tim


Tim

Peter N
11-08-2009, 04:48 AM
A very light film/dusting of spray paint or something similar?
Completely unscientific and probably totally unmeasurable, and certainly wildly random as to the amount of deposition, but will quite probably give you a tenth or two.

Peter

Black_Moons
11-08-2009, 05:19 AM
A single spec of dirt. Id suggest just breathing on the surface and waving your arms in the general direction. that should add a tenth or two..

On a more pratical scale, How about teflon tape, thats pertty thin..

Evan
11-08-2009, 05:55 AM
I did exactly that to my SB9 Paul to reduce the tendency to turn concave on a facing cut. It needed about .001" brass shim stock on the back Vee way to bring it into line. It made so little difference to the height that there was no reason the alter the tailstock. If you take the .001 x .707 = .0007 /2 = .00035 that is how much it raised the centre line. Raising the centre line by .0003 has no measurable effect on anything.

BTW, Glad plastic Cling Wrap is .0005 thick and is pure high density polyethelene.

Sparky_NY
11-08-2009, 08:51 AM
I recall that metallic wrap they use for candy bars and similar is extremely thin. Don't have any to measure right now.

Evan
11-08-2009, 09:22 AM
The old foil covered paper which is no longer used could be peeled apart. The foil was only about .0002" thick. I haven't seen any in years.

Glenn Wegman
11-08-2009, 09:37 AM
There is shim paper specifically for that.

I have some and it came in a set of three rolls, I believe .00025", .0005". and .00075".

Just can't recall where I got it, but I believe from a jig gringing guy!

Lew Hartswick
11-08-2009, 09:58 AM
The old foil covered paper which is no longer used could be peeled apart. The foil was only about .0002" thick. I haven't seen any in years.
Aw yes the way chewing gum came when I was kid. I remember
pealing that off to make a "tinfoil" ball. Gawd that has been a while.
...lew...

BillC
11-08-2009, 11:02 AM
What if you used JB Weld two small spots tighten the head stock down till its where you want it then let it set up and cure????. Bill.

Forrest Addy
11-08-2009, 11:12 AM
Criminy, scrape it. You could do it in two light cuts and improve the bearing in the process.

Mcruff
11-08-2009, 12:28 PM
The old foil covered paper which is no longer used could be peeled apart. The foil was only about .0002" thick. I haven't seen any in years.
We used to use this alot, but like Evan said I haven't seen any in at least 15 years. This stuff used to be actually sold in some machine catalogs for shimming purpose's, when you bougth it it was about 1/8" thick and about 1/4" x1" , peeled off what you needed.
The clear cellophane on a cigarette pack used to be thin (.0003), I have used that a bunch where heat is not involved.

Ryobiguy
11-08-2009, 02:15 PM
Ah yes, the SB headstock shimming. I tried this, went through a series of attempts on my old 13":
-shim up one side (or two,)
-take a cut several inches long
-run the test dial indicator on both the tool-facing side of the work, and then over the TOP of the work.

This way you see which direction the spindle is off-axis compared to the axis of the cutting tool.

I found it was a little tricky to get right, and I never quite finished at near perfection, but it was a big improvement since it was REALLY off before I started (something like .010 taper in 6"!)

The hard part was getting consistent results - it seemed that everytime I lifted the headstock, changed the shim (or didn't!), then locked down the headstock in position, the spindle axis moved a bit differently than what was expected.
Probably I was breathing and waving my arms too much for that extra tenth or two. :)

Anyways, one thought I had on how to get the partial thou shims -- if you can get some shims that are say, 1.5 thou, then maybe you could shim the other legs with 1 thou shim stock to get the partial thou difference on one of them. (Well, it's not that simple because of the V ways.)

-Matt

lynnl
11-08-2009, 03:10 PM
A very light film/dusting of spray paint or something similar?
Completely unscientific and probably totally unmeasurable, and certainly wildly random as to the amount of deposition, but will quite probably give you a tenth or two.

Peter

Another thought along that line: Possibly lay down a few thin coats (maybe a metallic paint) on something like wax paper, so that it could be peeled off after drying. Just a thought.

Evan
11-08-2009, 03:56 PM
I know this may go against the grain so to speak but an idea just occured to me. Apply a very thin layer of grease to the surface to be shimmed. Then sift on a grit of a known size such as a particular mesh of silicon carbide or even sand. Blow off excess with some air. Bingo, instant known shim that will also prevent relative motion.

JCHannum
11-08-2009, 04:14 PM
Shimming and tenths do not go together. Forrest is correct, if you need to work to that level of accuracy, scrape it and be done.

A simple difference in the amount of torque applied when tightening the assembly down can cause a tenth or so deflection. Aluminum foil, plastic wrap and other found materials will not necessarily be uniform and most all will have a squish factor to contend with.

Forrest Addy
11-08-2009, 04:36 PM
Thank, you JC.

You in the machinist trade should all know by now that there are times and places for expedients and a headstock re-alignment is not one of them. A headstock is fitted to the bed metal to metal in accurate alignment at the factory. Some small lower end engine lathe headstocks have a setscrew adjustment on the change gear end to shift the spindle axis slightly. The rest are scraped for spindle axis alignment with the bed.

A human hair is typically 0.003" in dia. Slice it 30 times like brad for a sub sandwich and each one of the slices will be 0.0001". A very small amount of material indeed; well within a simple blue and scrape. A SB 13 headstock is small and light. If the operator is muscular he won't even need hoisting gear for repeat fittings. No it's not simple and quick but it is permanent and accurate if done correctly and it will never ever shift.

Shim a headstock for alignment! Just the thought of such a shade tree method gives me a pain where I sit.

Evan
11-08-2009, 05:29 PM
I shimmed mine about 5 years ago and it hasn't moved since. It is very easy to do with a South Bend because of the way it is clamped to the ways in the same manner as the tailstock. It also involves no permanent modification and doesn't present the possibility of permanently screwing up the lathe. It can be reversed in a couple of minutes, something that cannot be said for scraping.

Doc Nickel
11-08-2009, 05:40 PM
Apply a very thin layer of grease to the surface to be shimmed. Then sift on a grit of a known size such as a particular mesh of silicon carbide or even sand.

-The grit will embed, giving variable results.

To the OP: since the value is so fine, why not just stone the high points?

Doc.

lane
11-08-2009, 06:02 PM
May be i`m off in left field some where.
But a South Bend head stock sits on a v and a flat way and is bolted down front and back with a large bolt . It does not slide are move at all . So how can it be out of alignment from the day it was fitted . If you have a alignment problem it has to be something else. If it moved around I could see ware being a problem . But is aligns to the v way and tighten down solid . So how can it be out of whack Un less you took it off and got some trash under it . Just my thinking. . My thinking is you are looking in the wrong place for the problem.

Evan
11-08-2009, 06:06 PM
Like many lathes the spec for the headstock when facing is +zero from flat and -something concave. Since it isn't wise to shoot for the edge of a tolerance band they usually face slightly concave. A very thin shim can bring it very close to flat.

Mcruff
11-08-2009, 06:14 PM
Shimming and tenths do not go together. Forrest is correct, if you need to work to that level of accuracy, scrape it and be done.

Well I have never scraped or seen a plastic injection mold scraped however I have in 28 years of building them seen hundreds of them shimmed (in tenths) and some of them are clamped up in presses that range in the 1000's of tons of clamping pressure. We have 3 large Husky presses at work and the smallest of them is 2,500 ton, the 2 large ones are 3,650 ton machines.
Scraping definitely has its place but don't ever discount what you can shim even in tenths.

ammcoman2
11-08-2009, 06:27 PM
I just measured some foil wrap on a bar of chocolate - yes, I just opened one up as I have to get my chocolate fix! As near as I can tell it is between 0.0002" and 0.0003" thick. This was from a large bar (300grm)of Swiss chocolate.

Geoff
(the fix is doing its work!)

GadgetBuilder
11-08-2009, 07:42 PM
Like many lathes the spec for the headstock when facing is +zero from flat and -something concave. Since it isn't wise to shoot for the edge of a tolerance band they usually face slightly concave. A very thin shim can bring it very close to flat.

My understanding of lathe alignment is that if the spindle isn't aligned to the ways then the lathe will cut a taper. Once the spindle is parallel to the ways then to adjust flatness/concavity while facing one must adjust the cross slide travel so it is perpendicular to the spindle axis or travels slightly toward the headstock as the CS advances, typically less than 2 tenths per inch.

I don't see how shimming the headstock can correct facing concavity without changing taper?

John

lynnl
11-08-2009, 07:45 PM
For a purpose such as is sought here, where a bearing surface is not the objective, would sanding be an acceptable alternative to scraping?

Yeah, I know it takes a lot of sanding to remove even a tenth, but faced with this problem I'd probably feel more comfortable trying to sand the area down while hoping to retain the original plane. (...or parallel to the original)

In any case, I like Lane's logic on the matter. That's what would scare me, trying to fix what ain't broke. :) I've done that all too often.

Mcruff
11-08-2009, 07:54 PM
For a purpose such as is sought here, where a bearing surface is not the objective, would sanding be an acceptable alternative to scraping?

Yeah, I know it takes a lot of sanding to remove even a tenth,
You'd be shocked at what you can polish and sand off in a matter of a few minutes. Heck with good paper on a soft surface you can take .0002-.0004 off in a few minutes over say a 1/2" square.

airsmith282
11-08-2009, 08:02 PM
It is possible to use tin foil once you put in in the you tighten the head back down it will likey goto half its thickness,might even go thinner. you never know till you try,

lane
11-08-2009, 08:16 PM
Besides again a lathe is not suppose to face dead flat , but concaved so to faced parts will sit to gether with out rocking . I say again . nothing wrong with way head stock fits the bed ways . Can not ware . something else wrong . may be to much ware in cross slide are ways .

darryl
11-08-2009, 08:19 PM
Just 2 cents worth- if the spindle axis is precisely parallel to the ways, in theory it shouldn't turn a taper. For that to be the case, the ways must also be perfectly parallel. Leveling is supposed to achieve this.

Once that is true, or just assume it's true for the sake of the discussion, the crosslide must be precisely square to the ways in order for a facing cut to leave a perfectly flat surface. In practice it won't because the parameters change as the cutting tool makes its way towards the center point. The depth of cut might vary with the change in SFM of the cut, and the way the metal curves around the cutting edge. Closer to center, the metal goes in a tighter arc around the cutter.

I don't know enough to say whether the concavity is designed in for this reason, or for this plus another reason, which would be so that flanges can be made that will compress tightly to a surface when bolted down, rather than the outer edge lifting slightly when bolted down.

As a hobby machinist, I'd rather that the lathe cut perfectly flat when facing, and not tapered for turning a length. I could live with the possibility that I'd have to leave a very slightly raised outer lip on a facing cut if I needed the thing to bear on its outer circumference for the best stability when mounted.

All things being right with the carriage and the ways, the headstock should be adjusted so you don't turn a taper over a length. For a very rough 123 of how to adjust the lathe, step one is align the ways, step two is adjust the headstock, step three is set up the tailstock. I'm only saying this because it's futile to adjust the headstock if the carriage can't ride the ways in a straight line and without tilting over the full length of it's possible travel.

Since there is no lathe (that I know of anyway) where you can adjust the perpendicularity of the crosslide with the ways, you would probably be better off to live with it turning a slight concave on a facing cut, as long as it isn't turning a taper over a length.

An aside here- the lathe I'm building, or converting, will have an adjustable crosslide so I can set an angle other than perfect 90. This means I can never get a perfect facing cut unless I make sure it's adjusted just right, but I do get the ability to turn a taper over a longer length than by using the compound. This one won't have a compound.

Hmm. Depending on what happens at work, I might get the time to work on that again soon.

Evan
11-08-2009, 08:34 PM
The reason I shimmed mine was because I was flycutting parts a full nine inches wide using all the travel of the cross slide and the maximum swing of the lathe. Because of this the flycutter was cutting both on the front side and the back swing each rotation. With the lathe aligned to cut a slight concave face it meant that the flycutter was cutting deeper on the near side compared to the back side leaving a ridge at the limit of travel near the middle.

Example of the setup:

http://metalshopborealis.ca/pics/34face3.jpg

http://metalshopborealis.ca/pics/34face2.jpg

I also needed the parts to be as square as possible.

http://metalshopborealis.ca/pics/34face1.jpg

airsmith282
11-08-2009, 08:38 PM
iam woundering if the spindel bearings could be a problem im sure like anything else, they also ware out and would cause a shift that can not be corrected unless you replace the spindel bearings,, just a thought

Evan
11-08-2009, 08:47 PM
There is nothing wrong with the lathe. The shimming is to bring the facing cut to as flat as possible without being convex.

darryl
11-08-2009, 08:47 PM
Bad spindle bearings would of course be a problem. After all, they are what holds the spindle in alignment with the headstock.

As I work with the lathe, I often think about how far out the workpiece actually is from the front bearing, even if it's close to the chuck jaws. It's easy to see that an error in a bearing will be magnified by at least double by the time you measure it at the workpiece.

JCHannum
11-08-2009, 08:56 PM
Lathes are designed to not face flat on purpose. They are designed to turn parallel to the length of the bed. Convex or concavity of facing is a function of the cross slide alignment, not the headstock. Start cranking the headstock around and you will start turning a taper.

darryl
11-08-2009, 08:59 PM
I'm reminded again about the Unimat lathe with it's adjustable headstock. Though it's difficult to get precisely parallel without some testing and tweaking, sometimes it's so handy. There is a pin to insert, but that gives you what- 89.5 degrees, or 90.03 degrees-

How about a variation on a rotary table to interface between the ways and the headstock? Would give an increase in turnable diameter as well, though requiring a change in the tool holder-

It would give an effective compound travel equal to the length of the carriage travel. Long thin tapers, anyone-

psomero
11-08-2009, 09:08 PM
we have some pretty thin stainless and brass shim stock at work. i used it to shim the column on my mill and it worked out great.

i'll try to snag a couple little pieces next time i'm there and if you PM me an address, i'll drop it in the mail...

Evan
11-08-2009, 09:12 PM
Lathes are not designed to face flat on purpose. They are designed to turn palallel to the length of the bed. Convex or concavity of facing is a function of the cross slide alignment, not the headstock. Start cranking the headstock around and you will start turning a taper.


The carriage alignment isn't adjustable. The headstock is. Shimming the headstock to face flat brings the right end of a long part closer to the tool. In theory this might result in turning a taper but in practice it actually compensates for tool forces that have a greater effect the farther the tool is from the headstock. The amount of change is very small, perhaps .001 in 8 inches. Normally anything that extended would be supported by the tailstock and that entirely eliminates the taper issue.

J Tiers
11-08-2009, 09:33 PM
If you can't get really thin shims, then see if you can get thicker ones in small increments......

For instance, a 0.0012 and a 0.0015 have a 0.0003 difference..... Shim all 4 spots, but use the thicker only where you need the small extra shim thickness, and the thinner elsewhere.

Sure, it moves everything, but only a tiny amount, and its the difference you want.



You'd be shocked at what you can polish and sand off in a matter of a few minutes. Heck with good paper on a soft surface you can take .0002-.0004 off in a few minutes over say a 1/2" square.

I've worked long and hard taking off a few tenths with sandpaper, rubbing a hardened part on it. The part was a bit of tube-shaped metal that I wanted 1/2" long, exactly. 0.3 OD, and about 0.125 ID. it took a long time.

Worked a long time to take off a few tenths all over a part scraping, too. Had lots of scrapings , nice pile, but hadn't taken squat off the part yet. And that was bearing down with a carbide scraper, with swarf flying.

Black_Moons
11-08-2009, 09:39 PM
Evan: you never cease to amaze and scare me. I was going to ask 'why on earth would you do that in a lathe' untill I realised the length of that part would never fit in a mill, and side milling it would be problematic at best with that much depth.

I do believe I read somewhere that lathes are indeed designed to face concave so that two faced parts will 'mate' on the outside insted of roll about on an inside high spot. Usally something like a few tenths per inch but im not totaly sure on that number.

as for turning taper, if your headstock was mis aligned slightly, would'nt twisting the bed correct for that? (Considering the twist is a degree or two and ignoreing any fraction of a thou tool lowering by the twist being non consistant with diffrent work diamiters)

Leaving me to hope the OP is only aligning the headstock for facing cuts only.
Also making me wonder exactly how concave the OP's facing cuts are?

Does the OP really need precision facing cuts on all future large parts under power feed?

Or could the OP get away with making the current part using the compound to face, adjusted to be 'perfect' flat with a TDI and some messurement surfaces/tools.

Might there be something wrong with the OP's facing cut messurement proceedure? Like the fact that SFM constantly falls in a facing proceedure without a VFD to incress RPM's and hence the cutting forces are not constant, Or his test workpeice may of warped due to stress/heating/cooling/removing from the chuck, Or messureing the flatness with a flawed method, or dirt interferance beween mating surfaces used for messurement.

Does the facing produce a convex surface if you do it starting from behind the part?

JCHannum
11-08-2009, 09:49 PM
The carriage alignment isn't adjustable. The headstock is. Shimming the headstock to face flat brings the right end of a long part closer to the tool. In theory this might result in turning a taper but in practice it actually compensates for tool forces that have a greater effect the farther the tool is from the headstock. The amount of change is very small, perhaps .001 in 8 inches. Normally anything that extended would be supported by the tailstock and that entirely eliminates the taper issue.

BS; If the headstock is not parallel to the lathe bed, the lathe will cut a taper.

boslab
11-08-2009, 09:54 PM
i dont work to the precision you guys are at, honest, i'm happy if two parts fit together properly, never mind having the urge to shim tenths, i have to say at this point i'm dammed impressed, i think i'd be challenged to shim to ten thou!
Seriously if you need that kind of accuricy a new machine would be a good start i think?, then tweak it but it seems like the quest for precision is getting the better of some folk [certainly no offence intended, i'm jealous with my eysight these days if i didnt have digital measuring kit and had to use a vernieer mic i'd be doomed]
most of the machines we get to have in a home workshop dont seem to be the kind that could hold these extrime levels of accuricy, or at least repeat them come on guys +/- 1/64" or is that too tight lol [it is for me]
i think i know of a krypton absolute interferometer going cheap, collect youself needs new bulb!"
kind regards
mark

Evan
11-08-2009, 10:28 PM
BS; If the headstock is not parallel to the lathe bed, the lathe will cut a taper.


Mine doesn't and the bed isn't twisted either. The carriage design on the South Bend is an overconstrained design. The front and rear corners ride on vee ways which makes it very difficult to rescrape after it leaves the factory.

The cross slide must be aligned so that it runs perpendicular to the ways and carriage travel, not the spindle axis. To do otherwise will introduce an error that must be compensated by introducing another error in the position of the tail stock if one wishes to turn faces at right angles to the diameter. That is unacceptable. By cranking the headstock slightly to produce slightly concave facing in a chuck no adverse effect is produced on the primary high accuracy mode of operation which is to turn between centres. If the tailstock is aligned to the point of a center turned in place in the spindle then it will share correct alignment with the turning centre point of the spindle and the carriage will be perpendicular to the turning axis. That is the only way that the lathe can be set up correctly.

Cranking the headstock to face flat only requires that the tailstock be adjusted to match. That brings between centres alignment back to parallelism with the ways.

oldtiffie
11-08-2009, 10:55 PM
I am surprised that no one has even queried whether the cross-slide is excessively worn, not straight or even whether it is at right angles to the path that the carriage/saddle tracks/moves in/on the presumably worn lathe bed (axis).

For what its worth, given that the OP referred to working "minutes" (of arc) that one minute of arc is 1/60 degree the tangent of which = tan 1/60 = 2.9 (say 3) x 10^-4 ~ 0.0003"/inch ("three tenths" per inch) which if applied to the face/s of a 90 degree vee-slide/way on a lathe would result in both a vertical and lateral movement of the head-stock of sq.rt 2 = 2^0.5 = 1.414 x the thickness of the shim/s or about 4.5 x 10^4 ~ 0.00045"/" = 4.5 "tenths" per inch.

I have not allowed for shim size error or compression.

Evan
11-08-2009, 11:01 PM
The vertical movement will be half that because it is centred between the ways. No lateral movement will result as it will be shimmed equally on opposite sides of the rear vee way at opposite ends of the headstock.

Jim Hubbell
11-09-2009, 12:21 AM
I keep a "run capacitor " core handy for the few times I might need a shim measured in tenths. It is tough and cheap.


Jim

Forrest Addy
11-09-2009, 12:33 AM
There have been a number of standards developed over the last century that deal with the permissible amount and direction of machine tool alignments at acceptance trials of the machine by the customer. I have some recollection of them pertaining to small lathes. In an engine lathe the spindle axis in the horizontal plane is set parallel to the carriage motion within 0.0002" per ft as determined by a dial test indicator and a calibrated proof bar (the two callar test is in my opinion more reliable.) The spindle axis in the vertical plane shall be parallel to or rise 0.0005" per ft from the saddle motion. The cross slide axis shall be perpendicular in the horizontal axis (as in a radial plane) to 0.0005" per foot acute so a perfect tool would machine a convex if there was to be any error. That is if I recall correctly. I don't think I am off by far in he numbers.

These accuracies are developed at the factory by hand scraping in days of yore or by way grinders in the present day in near clean room conditions. Regardless even lower tier import engine lathes meet these requirements. The fitting and assembly of machine tools in a production setting goes quickly and the end product is warranted to the manufacturer's published specs.

Machined accuracy tests are as best difficult to achieve and at worst unreliable. Machining tests intended to demonstrate errors of 0.0002" per ft are often confounded by tool wear, build up in the cutting edge, temperature etc. If a machine will turn and bore with consistant but cancelling error using the same tool geometry that's about the best you can hope for. If the machine will face flat feeding outand face concave feeding outthat's about the best you can hope for.

For that reason machining tests in machine tool runoff are intended to demonstrate stock removal, reliability of the axis feeds, accuracy of the measurement system, etc not the overall geometrical accuracy and linearity of the axis systems.

Machine tools with time on them may be off depending on the amount and distribution of the wear. Factory specs are in this case only something to dream about and maybe used for guidance for interim correction of wear problems should full reconditioning be out of the question.

(edited to correct mistatement: I wrote convex instead of concave. There is a difference that even I can eventually understand convex makes a bad coffee cup but a good bosum. OK, got it. I think.

macona
11-09-2009, 01:23 AM
I did exactly that to my SB9 Paul to reduce the tendency to turn concave on a facing cut. It needed about .001" brass shim stock on the back Vee way to bring it into line. It made so little difference to the height that there was no reason the alter the tailstock. If you take the .001 x .707 = .0007 /2 = .00035 that is how much it raised the centre line. Raising the centre line by .0003 has no measurable effect on anything.

BTW, Glad plastic Cling Wrap is .0005 thick and is pure high density polyethelene.


Apparently lathes are supposed to cut slightly concave. Didnt know this until I was rebuilding the carriage on my 10EE. A.S.A. Standard is .0005 concave.

OOPS! too slow. Forrest beat me to it.

Heres the checkout sheet for a monarch. Sorry for the quality. I didnt scan it.

http://i65.photobucket.com/albums/h228/macona/monarchee.jpg

JCHannum
11-09-2009, 01:46 AM
Listen to Forrest folks. The spindle axis is parallel to the carriage motion to 0.0002" per foot travel. If exact perpendicularity to the spindle axis is required, the compound is used. That is what it was designed for.

The set up and alignment of a lathe is a very well defined procedure. The Connelly book is probably one of the best sources for that information. It describes what is a proper setup for the lathe and the methods involved in measurement and achieving that goal. It also prescribes the order in which these steps must be done so as not to negate previous efforts.

Jumping into the middle to correct some perceived or imagined "error" will lead to nothing but grief.

Paul Alciatore
11-09-2009, 03:40 AM
I just measured some foil wrap on a bar of chocolate - yes, I just opened one up as I have to get my chocolate fix! As near as I can tell it is between 0.0002" and 0.0003" thick. This was from a large bar (300grm)of Swiss chocolate.

Geoff
(the fix is doing its work!)


What Brand?????

Paul Alciatore
11-09-2009, 03:49 AM
Lathes are designed to not face flat on purpose. They are designed to turn parallel to the length of the bed. Convex or concavity of facing is a function of the cross slide alignment, not the headstock. Start cranking the headstock around and you will start turning a taper.


I am turning a taper. About 0.003" to 0.004" on the diameter over about 6". That's what I am trying to cure, not a facing cut. I would assume the cross slide, not the headstock, would be the source of a concave facing cut and that this is deliberate.

I suspect my SB was assembled, not by the factory, from parts that were not scraped to match. I am not overly sure of the cure and shimming would tell me if it is the headstock or not. Only after I am sure, would I presume to scrap or sand it. Shiming seems to be a reversable way of doing this.

Paul Alciatore
11-09-2009, 03:59 AM
If you can't get really thin shims, then see if you can get thicker ones in small increments......

For instance, a 0.0012 and a 0.0015 have a 0.0003 difference..... Shim all 4 spots, but use the thicker only where you need the small extra shim thickness, and the thinner elsewhere.

Sure, it moves everything, but only a tiny amount, and its the difference you want.

......

Yes, I am aware of using differences and have done so on other shiming jobs. But where do I get such differences? Most shim stock I see is 0.001", 0.002", 0.003", etc. They don't seem to do small differences.

Evan
11-09-2009, 04:09 AM
Paul,

Have you placed a sensitive level on the carriage and then watched it as you crank the carriage from one end to the other? It is the change that matters, not the absolute level so the level itself may be shimmed to make it readable.

Example:

http://metalshopborealis.ca/pics3/level1.jpg

http://metalshopborealis.ca/pics3/level2.jpg

Black_Moons
11-09-2009, 04:26 AM
If your lathe is turning a taper id highly recommend looking at 'leveling' proceedure, Or more to the point, untwisting the bed itself.

Basicly you shim *under* one side of the bed of the lathe itself (or using adjustable feet under the lathe stand) and the entire bed and ways of the lathe twists.

You may not believe me that your 1000lb chunk of iron twists like a wet noodle depending on your leveling feet, but it does.

Once its level, you actualy turn parts and adjust the twist on the bed untill it turns without taper. You do NOT adjust the headstock. Be VERY thankful you did'nt adjust the alignment of your headstock and it has V ways and please stop trying to or you'll hate yourself.

Evan
11-09-2009, 04:54 AM
The bedways should only be adjusted by shimming the feet of the lathe so that they are coplanar. No attempt to correct any other condition should be made by twisting the bed away from a coplanar condition. The bedways are the reference for everything else so if they aren't straight and lying in the same plane then no other adjustments are meaningful.

Once that is done the alignment of the headstock can be accurately checked optically. Make an adapter to hold a riflescope coaxially on the left end of the spindle so that it can be used as a boresight. Place a target on the far end of the lathe to the right. Align the reticle using the elevation and windage screws so that the point of aim does not change when the spindle is rotated.

The scope will now be exactly aligned with the spindle axis. Put the tailstock in place and install a dead centre. Use the scope to verify the location of the point of the centre relative the spindle axis via the crosshairs. Move the tailstock along the ways and watch for any deviation in alignment. This will quantify wear of the ways and any degree of misalignment that may exist in the headstock to ways.

Forrest Addy
11-09-2009, 05:04 AM
Evan, you have a telescopic sight that close focuses to a couple of feet?

Black_Moons
11-09-2009, 05:37 AM
Evan: I don't mean twist the bed outta alignment, I mean do proper test cuts for fine tuning the alignment as the level method isent 100% precise, its the actual cut it takes that really matters. (And even then MUCH care must be taken for the test, like using the proper material to test cut and super sharp HSS bits, and locking all the other axises)

Evan
11-09-2009, 05:54 AM
Certainly. I have a cheap little Weaver, the type you would find on a kid's first 22. It will focus to about 1.5 feet and can easily pick out a single pixel on the monitor screen. The monitor displays 92 pixels per inch so that is an accuracy of about .011". That will certainly be enough to check coarse alignment on a lathe headstock. More importantly, you can detect changes smaller than that as the position of the image changes in the field of view.

This isn't the best picture since I don't have an adapter for the camera to the scope. I had to hold it to the eyepiece to take the picture.

http://ixian.ca/pics6/weaver.jpg

Black_Moons
11-09-2009, 06:09 AM
Look out forrest, evan is gunning for ya!

Evan
11-09-2009, 06:24 AM
Evan: I don't mean twist the bed outta alignment, I mean do proper test cuts for fine tuning the alignment as the level method isent 100% precise, its the actual cut it takes that really matters. (And even then MUCH care must be taken for the test, like using the proper material to test cut and super sharp HSS bits, and locking all the other axises)


Altering the alignment of the bed to suit the aligment of the spindle axis in order to make a non tapered cut means you are tuning the bed to the spindle. That is backward. The reference is the bed. You align the bed so it is coplanar and then everything is referenced to it.

The accuracy of the bed alignment depends in part on the tools at your disposal. I have a very accurate master precision level and it is more than good enough to align a lathe bed. There are other ways to accomplish the same thing without such an expensive tool. Turning an ordinary bubble level partway on it's side will make it much more sensitive. All that is needed is a relative indication. It's the change that matters, not the absolute value.

In fact, if you have an Apple I-Phone there are several freeware applications that use the internal accelerometer to report relative level with an accuracy of 1/100 degree. The accelerometer is actually more accurate than that but that is the best the software will report.

http://iphone.wareseeker.com/free-great-level/

ammcoman2
11-09-2009, 08:28 AM
What Brand?????

It is President's Choice sold by Loblaws (a grocery chain) here in Canada. The chocolate is "Dark Chocolate with Almonds" and is made in France.

I could send you the wrap from the next bar that is currently in "inventory" if you like. Should be opened in about 3-4 days as we try to limit consumption. The fear of becoming a chocoholic is a real one!

Geoff

JCHannum
11-09-2009, 08:48 AM
I will stick with my previous recommendation of listening to Forrest and getting the Connelly book if one is truly interested in learning the proper methods of lathe alignment.


I am turning a taper. About 0.003" to 0.004" on the diameter over about 6". That's what I am trying to cure, not a facing cut. I would assume the cross slide, not the headstock, would be the source of a concave facing cut and that this is deliberate.

I suspect my SB was assembled, not by the factory, from parts that were not scraped to match. I am not overly sure of the cure and shimming would tell me if it is the headstock or not. Only after I am sure, would I presume to scrap or sand it. Shiming seems to be a reversable way of doing this.

You are correct that concavity is a function of the cross slide, not the headstock.

I will not pretend to be an expert, but have absorbed some basic information from Forrest and others who do know what they are doing.

Even if the headstock is from another machine, differences in machining would more likely result in height or displacement from center and not be angular in nature. The V-ways in the headstock and the spindle bore must be parallel from the start and the manufacturing sequence is such to insure this is the case. This relationship cannot be compensated for by other means, and it is a fairly safe assumption that the source of taper is not there.

Forrest might be good enough to add his wisdom here, but my first approach would be to remove the headstock and scrupulously clean all mating surfaces. Then lightly stone those surfaces. Inspection after stoning will show any remaining high spots that might have occurred over the years. Additional stoning or light scraping might be needed to remove these.

Clean again and reassemble. I would guess you have already done this, but it does not hurt to recheck. I would also assume that the bed has been leveled with the most accurate means available. Retest to determine the amount of taper if any. I do necessarily recommend Rollie's Dad's Method of lathe alignment, it can be searched on the net, but at times, it or similar means are often needed for final alignment of a lathe once it is properly set up. As I recall, this is covered in the Connelly book. (I no longer have a copy at my disposal.)

Mcgyver
11-09-2009, 09:08 AM
For a purpose such as is sought here, where a bearing surface is not the objective, would sanding be an acceptable alternative to scraping?
.

Lynn, imo, no way....because there is very little control over how much is removed where. you easily reduce the bearing area between the two parts



I will stick with my previous recommendation of listening to Forrest and getting the Connelly book if one is truly interested in learning the proper methods of lathe alignment.

Very sound advice. shimming is not the way to do this job. they can move & squish but most importantly you do NOT get the level of control or granularity that you do with scraping. its quite easy to scrape a surface (light finish scraping) and not move the needle of a tenths indicator...in other words the depth of cut can be so fine that you have the control to nudge things a tenth one way or the other

....ok, but once stated, who cares if others dissent....you know some will waste a 100 pages of time on it, soooo moving on....

the question now is how to scrape this, I'd like to here some the experienced guys like Forrest explain how the axis of a high quality lathe like a monarch is scraped so that it will hold a tenth over some distance.

Is it as simply as getting bed spot on (levelled perfectly, no wear)....placing the headstock on and then with the carriage indicating a precise ground bar? Seems to me there has to be more to it; there's error in the bar, error in how the bar is held, etc. To get the HS axis to within a tenth over some distance would i think require resolution finer than a tenth.

...or do you put it all back together and turn a test piece, measure it with a indicating mic etc. that would overcome error in work holding, but limits the sample distance to being quite short. That sounds like a painful month of Sundays if it takes ten iterations!

So what are the options? how are the high quality lathes aligned?

rowbare
11-09-2009, 09:23 AM
The old foil covered paper which is no longer used could be peeled apart. The foil was only about .0002" thick. I haven't seen any in years.
Many craft store carry metal leaf. The most common are gold, silver and imitation gold. It would probably do the job. Comes in packages of several sheets for a few dollars. I have seen in at Michaels and others.

http://en.wikipedia.org/wiki/Metal_leaf

Paul Alciatore
11-09-2009, 11:00 AM
With regard to leveling the lathe, I had done that previously when it was on a good, solid concrete floor. However, it is presently mounted in a trailer and any such effort would be totally wasted. It is on a steel table and for the present that is the best I can do. I did check it with my best level when I put it in there using Evan's idea of the level on the cross slide and some shims to make it readable. It was close at that time, which was about a year and a half ago. I will recheck it again using the same procedure.

But basically I am stuck with a bad situation and am trying to make the best of it. The lack of a solid base is another reason why I would not even dream of scraping anything at this time. I would likely do more damage than good. It just seems to me that I can do better, even with such a poor mounting. I am not going for tenths in the parts I am turning, just in the shimming process. I would hope that by using shims I could get a six inch part down to about one thousanth variation instead of thre or four.

I do appreciate his advise and I know it is good, but with apologies to Forrest, but I am NOT going to scrap or sand or otherwise remove metal at this point.

Evan
11-09-2009, 11:05 AM
the question now is how to scrape this, I'd like to here some the experienced guys like Forrest explain how the axis of a high quality lathe like a monarch is scraped so that it will hold a tenth over some distance.


Unless the Monarch has the same bedway design comparisons aren't useful. The bedway design of the South Bend is a double Vee way. This provides very positive location of the components but is what is called over constrained. A kinematically constrained design using parallel Vee ways would have the carriage and headstock bearing on three points, not four. With a four point bearing system the problem of accurately fitting the mating parts is greatly increased. Changing that relationship is even more difficult. To scrape it correctly material will have to be removed from as many as 16 surface areas, eight of them critical. To shim it material must be added to as many as eight surface locations all of them critical but easily adjusted. Because the headstock will be slightly raised instead of lowered the other surfaces don't present an issue.

Shimming isn't some sort of poor cousin in machine design. It is a fully accepted method of adjusting part locations and orientations. There isn't a machine tool made that doesn't have shims in it somewhere. Gibs are a prime example.

I would suggest that Jim read Precision Machine Design by Alexander Slocum of MIT. While they were still available back when I designed my milling machine I downloaded the entire series of open courseware files from his Engineering Design course. It's 21 Chapters of his lectures and course notes with illustrations and discusses all of the pertinent elements of how and why machines are designed and how and why errors are dealt with. Much of it is new material that departs from the "old ways" as it takes advantage of techniques for analysis that didn't exist not so many years ago.

Unfortunately the files are no longer available because they are being used as the source material for another book by Professor Slocum.

Forrest Addy
11-09-2009, 11:33 AM
The whole of the business of aligning a machine tool is to verify that the bed, baseplate, base casting, etc whatever is in the plane originally set at the factory from which all other axes of rotation and linear motion are established. Any attempt to make corrections without re-stablishing the original references may lead to endless errors, misleading results, and frustration.

If the lathe is on a trailer it can still be aligned (leveled) with solid screw jacks mounted on concrete pads. The jacks are positioned to roughly level the trailers and to take most of the spring out of the structure. Additional jacks are positions under the trailer directly under the points of support of the machine tool addressed. These screw jacks don't have to be fancy. A large diameter bolt tacked by the head to a steel pipe tripod and fitted with a coupling nut will work as well as one of Armstrong's finest.

As was alluded to earlier the machine does not have to be "level" but the twist, hump and sag has to be adjusted out so the bed or whatever lies in the factory plane. A precision level happens to be the simplest tool to use in establishing a true plane althogh optics, lasers, water levels, piano wire, etc has been used to great success over the years. I happen to have a wire sag table in Exel for 0.010" music wire under 22 lb tension if anyone wants a copy. If I can find it.

Parenthetically, a level surface is not straight. It's actually a spherical surface having the same radius as the surface of the Earth). A level line is a perpendicular to a line radiating from the Earth's center of gravity. It is not straight but has a roughly 4000 mile radius. Anyone care to work out the height of a 100 foot chord having a 4000 mile radius? It aint much.

An earlier comment about ensuring the fitted surfaces between bed and headstock have to be clean, the burrs removed, and assembled with a little oil or grease (to prevent air intrusion) is right on. Assessments of spindle alignment are worthless unless this connection is clean and metal to metal to begin with. And yes, final fitting and verification of many machine tool axis elements is frequently accomplished with tiresome partial assembly and disassembly.

If you think this is a PITA on a small engine lathe headstock imagine the hassle when it involves the headstock (weighing many tons) of a horizontal boring mill. When assembled on its column the spindle has to tram in X and Y to 0.0002" per foot or better. This can be quite a carnival involving the machine tool rebuilder, his helper, the rigger and his helper, and the crane and its operator - plus white hats and tourists subconsciously hoping for a bloody disaster. Better have your ducks in a row when your every action is being bird-dogged by upper management.

JCHannum
11-09-2009, 12:45 PM
Thanks for the words of wisdom Forrest, decades of experience will trump reading a book any time.

If I recall, Paul's lathe is a 9" South Bend or a light Ten. In either case, if it has been moved, it should be set up properly from the ground up, installed on as solid a base as available and leveled as well as possible before any further steps are taken.

I recommend using the tried and true methods of leveling, in particular using the level directly on the ground, unworn surfaces of the lathe bed as is standard practice. Locating the level on the compound and traversing the bed with the carriage will be subject to error introduced by wear present in the ways.

Mcgyver
11-09-2009, 01:04 PM
The whole of the business of aligning a machine tool is to verify that the bed, baseplate, base casting, etc whatever is in the plane originally set at the factory from which all other axes of rotation and linear motion are established. Any attempt to make corrections without re-stablishing the original references may lead to endless errors, misleading results, and frustration.


I'd define it a little differently, we want to do a lot more than verify that its in the original plane set by the factory. If we are on this path its likley either because the factory settings weren't good enough or we've reground or scraped the bed.....in other words the trick is what to do once we've figured it isn't where it should be.

I'm versed with scraping V ways and what leveling accomplishes...was more wondering if there were better techniques than a test cut to measure how far out the alignment is are. I't would be in Connelly's book, just dont have a copy at work


the spindle has to tram in X and Y to 0.0002" per foot or better

can you explain how this is done?

Evan
11-09-2009, 03:36 PM
Thanks for the words of wisdom Forrest, decades of experience will trump reading a book any time.


Then I take it that I shouldn't bother finding Connelly's book?



I have had a PM from Forrest on the subject of shims. I will not reveal it but I will post some information on currently accepted design practice in that respect.

Keep in mind that Paul does not wish to alter the machine, period.



Shims? Make mine laminated!June 5, 2003

Laminated shims make life easier for designers and machine operators, as long as they're properly specified.

Chief Engineer
Spirol International
Danielson, Conn.


Laminated shims come in a variety of shapes and sizes.

Laminated shims can be peeled down, layer by layer, until they are the proper thickness.

Intercomposite is a proprietary Spirol material made of glycol polyethylene terephthalate film. It is the lightest and lowest-cost material listed.

It's often quicker and less expensive to use a shim, a thin piece of metal or composite, to level a machine tool or make components fit together than grinding and machining to make the same mechanical adjustments. Shims act like those folded pieces of paper people place under table legs to prevent the table from wobbling. They are also used to quell vibrations and silence noisy equipment. But laminated shims are more versatile and economical than slivers of paper or machined wedges.

Laminated shims

Laminated shims have peelable layers of metal or composite which are removed until the shim has the proper thickness. Adjusting them is as easy as peeling off laminations with a knife, or in the case of some materials, using no tool at all. They are built up from layers of precision-gauge metal foil or composites. Layers are bonded into a rigid structure that appears and functions as a solid sheet or plate. The finished shims withstand reasonable handling, including shearing and machining.

Laminated shims are produced by surface-bonding layers of precision-metal foils or composite films and resin adhesive into sheets. The foils can be aluminum, stainless steel, carbon steel, and brass, among others. Final shims range from 0.006 to 0.250 in. thick. The bond is made by heat and pressure, which hardens and reduces the resin to the point it is almost undetectable At temperatures above approximately 300F (150C), however, the bonding agent may deteriorate and there will be a negligible loss of total thickness. But even heating the resin to above 446F does not influence the performance of the shim.

Shim considerations

Designers should know the forces that will be placed on the laminated shim before choosing a shim material. And shim faces should not be exposed to friction-causing motions unless the shim has a PTFE treatment. Otherwise, the shim could delaminate. However, if parts have location holes, the only forces will be those exerted by the tightening of screws. In this case there are no limitations on using laminated shims.Laminated shims should be machined rather than stamped. Machining leaves clean edges that make shims easier to peel. Machining also prevents burrs from forming when peeling layers away. The burrs are a by-product of die rolls and stamping.

Laminated shims can be partially solid and partially laminated. This type of shim is either half-solid or three-quarter solid, depending on the ratio of solid section to total thickness. Standard thicknesses for the solid part are 0.062, 0.094, and 0.125 in. Semisolid shims are used to add rigidity to a design, accommodate a bearing surface on one side of the shim, meet requests for a thick shim that will have minimal adjustments made to it, and to lower costs.

There are some size restrictions on laminated shims. For example, brass laminated shims larger than 12-in. diameter and laminated shims larger than 20-in. diameter must be made in sections. As a general rule of thumb, wall thicknesses should never be less than three times the total material thickness. Edges of laminated shims should not be rounded, and deburring laminated shims may make them difficult to peel.

--------------------------------------------------------------------------------

Why use shims?

There are three basic reasons for designing shims into assemblies and devices:

Tolerance compensation. Shims eliminate the time and cost of putting precision tolerances on mating components. They also compensate for accumulated tolerances during assembly. It is better to design in a shim than to discover later you need one at the assembly stage.

Precision alignment. Shims align parallel and angular surfaces when interfacing elements must be coupled.

Wear compensation. Shims compensate for wear and are often designed to be the sacrificial component so the basic equipment retains its original accuracy.


http://machinedesign.com/article/shims-make-mine-laminated-0605

JCHannum
11-09-2009, 05:42 PM
I highly recommend the Connelly book to anyone with a desire for information on the proper methods of inspecting machine alignment and the sequence these should be done in. However, simply reading this or any book does not an expert make.

I have no objection to the use of shims. They have a definite place in machine building and assembly. There are also places where their use is not appropriate. Shimming the headstock on a South Bend lathe is one of those places.

Paul's decision is his to make, whether he chooses to shim or not is up to him. However it is doing a disservice to others reading these posts to give the impression that shimming is an acceptable solution to this problem. It is not, it is a band aid at best.

Is Tiffiepedia a communicable disease? Or is there some other reason for posting meaningless information to prove some non-issue here?

beanbag
11-09-2009, 05:47 PM
I may have missed this post, but how did the OP determine that the headstock was misaligned?

Evan
11-09-2009, 05:57 PM
Is Tiffiepedia a communicable disease? Or is there some other reason for posting meaningless information to prove some non-issue here?


I have no idea what you mean. Practical experience is very valuable of course. It teaches one how to do a job and do it correctly providing you are correctly taught. However, the one thing it frequently doesn't teach is why things are done in a particular manner. That is where textbooks are important.

BTW, exactly where do you think the information in a textbook or reference comes from?

If you think information on the use of shimming is meaningless then why are you posting in this thread? That is the topic here.


There are also places where their use is not appropriate. Shimming the headstock on a South Bend lathe is one of those places.


That is your opinion. It doesn't have any support in mechanical reality. There is no valid engineering reason that shims are inappropriate for this particular application.

JCHannum
11-09-2009, 05:58 PM
I may have missed this post, but how did the OP determine that the headstock was misaligned?

You need to read through a few posts to find it out. The lathe is cutting a taper, but he has not identified that the headstock is misaligned.

It is always best to identify the source of the problem before effecting a repair.

beanbag
11-09-2009, 06:11 PM
I haven't looked up the proper headstock aligning procedure, but what I did was put a smooth rod of constant diameter (shock absorber piston shaft) in a 4jaw and indicated it near the chuck and a few inches out. In retrospect, I should have indicated from both sides, but oh well.

To indicate the tailstock I would have used an idicator mounted on the chuck, accounting for indicator droop, of course ;)

JCHannum
11-09-2009, 06:11 PM
I have no idea what you mean. Practical experience is very valuable of course. It teaches one how to do a job and do it correctly providing you are correctly taught. However, the one thing it frequently doesn't teach is why things are done in a particular manner. That is where textbooks are important.

My Point exactly. To do this job properly, I will continue to advise the uninformed to get the book and listen to Forrest.

BTW, exactly where do you think the information in a textbook or reference comes from?

If you think information on the use of shimming is meaningless then why are you posting in this thread? That is the topic here.

Wrong again, the topic is using shims as a means of aligning the headstock on a South Bend lathe to correct taper turning.


.......................

John Stevenson
11-09-2009, 06:24 PM
A lathe will cut a taper if it's not correctly levelled.

If you do need to shim it to tenths then use the correct factory shims, just cut the picture of the heastock off an old advert and by the time it's been oiled and clamped it will be about 2 tenths.

.

Mcgyver
11-09-2009, 07:06 PM
Do so with shims is kind of like solving the problem with lego building blocks; they're way to big for the kind of fine dimension-to-suit adjustment required, but more importantly its stuffing a rectangle into a wedge shaped space.....ie you are shimming at one one or the other of the headstock to change the angle. Stuff a rectangle into a wedge shape void and the area of contact is quite small. Poor practice imo as rigidity demands maximizing the bearing points. I'd want the length of both head stock registers to be fully engaged with the lathe bed

darryl
11-09-2009, 09:40 PM
On my lathe, the vee in the headstock sits above the front vee way. There are setscrews there that bear against one side of the vee, and the headstock has flat bottom surfaces that sit on the flats of the bed. Some dubious looking bolts hold it down.

I haven't had a problem with alignment in the several years that I've had it. I only mention this because this is a way already in place to make the adjustment, and I haven't had it shift.

I wonder if there's a way on the SB to install setscrews for this purpose.

The other thing I was thinking is that there hasn't been any mention of whether the lathe WAS in alignment before it was moved- if so it could well be that the bed is twisted slightly and causing the taper. If that's the case, arranging to 'untwist' it would be the better method of increasing the accuracy. It's not the proper method to align the lathe, but if the headstock is aligned and the error is due to bed twist, then untwisting it IS the way to rectify it. Likely that will be short term and will need touching up whenever a precision need arises. Altering the headstock for these touch-ups would not be the thing to do.

My opinions only, of course.

Mcgyver
11-09-2009, 09:50 PM
The other thing I was thinking is that there hasn't been any mention of whether the lathe WAS in alignment before it was moved- if so it could well be that the bed is twisted slightly and causing the taper. If that's the case, arranging to 'untwist' it would be the better method of increasing the accuracy. It's not the proper method to align the lathe, but if the headstock is aligned and the error is due to bed twist, then untwisting it IS the way to rectify it. Likely that will be short term and will need touching up whenever a precision need arises. Altering the headstock for these touch-ups would not be the thing to do.
.

100% agree and that's a good reminder. Comments about material removal, at least mine, are somewhat theoretical in that they assume lathe ways and leveling are perfect and aren't the source of error. Like Jim pointed out, you have to be certain about whats wrong before you try to fix things.

Black_Moons
11-09-2009, 10:04 PM
One more vote for *align* the bed before precision turnings should be done, Leave the poor headstock alone.

In a trailer your in a mobile envorment like a ship, except you can actualy get level sometimes. If all its bolted to is a table, that table is likey NOT straight

You may think it is, but its not. The table is not straight, the trailer is not straight, Nothing is straight, Its all spaggeti noodles flooping around at the slightest weight inbalance and must be straightened.

On ships, some people have reported recording deflection on the lathe just from *walking* around the lathe. I doubt your trailer is that sturdy.

Infact its likey you will need to straighten the bed whenever the trailer is moved, and again whenever the utmost precision is required. Thankfuly if you setup leveling feet (Even UNDER the table its bolted to) the table itself will twist the lathe into shape when you alter those feet (just a few fractions of an inch, thats all thats needed), So theres no heavy lifting or disassembleing or shiming required. Doesnt that sound nicer?

I know its hard to believe that your nice expensive massive lathe is really a wet floopy noodle, and your solid steel table is really twisting and flexing, and your whole trailer, no matter how massive is likey bent way out of shape, but on a machinists scale of precision, its all true, And it will all affect your alignment and make you turn a taper. And is generaly to be fixed first long before the heatstock is ever looked at.

oldtiffie
11-09-2009, 10:25 PM
I think I recall the OP said that his lathe was set up in a trailer or a container. It will not be a lot different to setting up on a "springy" bench - or a lot worse. It may "move" as soon as he moves himself or a load or anything else in the vicinity of the lathe.

I haven't seen anything that I recall that said that the lathe was on a really solid base or a concrete slab - or what-ever.

Nothing said about wear on the bed/ways, under the saddle or in the cross-slide either - or what "slack" ("end float" and "lateral" movement) there is in the spindle and its bearings.

And what about the tail-stock?

How do we know if there is any vertical or horizontal "hogging" or "sagging" in the lathe bed or any significant "fixed" twist as opposed to elastic twist?

Having the lathe as good as you want may be one thing but having it as you need it may well be another easiest/best route to take.

J Tiers
11-09-2009, 11:00 PM
The problem of a lathe on a not-very-stable base is completely not solvable in any other way than by creating some sort of a stable base FIRST, and then working with the machine to "level" it or "untwist" itor whatever you want to call the process.

if you don't do that, you may as well just accept what you are getting now, because you won't do better for long.

Sure, shim it if you want to in that case. Lots better than scraping as it makes no permanent change.

However, long use of a lathe in a twisted condition will tend to wear the bed and carriage in a 'twist", which will be very nasty when or if it is later leveled.

Can't you re-inforce the floor and form a stable platform just for the machine?


Altering the alignment of the bed to suit the aligment of the spindle axis in order to make a non tapered cut means you are tuning the bed to the spindle. That is backward.



Which is one big reason that 'Rollies Dad's method" is not an appropriate substitute for a level, should you be so tempted...........

gmatov
11-10-2009, 02:18 AM
I gotta call BS on some of this. Even an SB 9 on a monolithic mount is not gonna twist as much as some of you are saying.

Them who are saying that a ship mounted lathe are gonna twist all over the place under weather are wrong. They don't bolt it to 3/8 skin of the ship.

I don't know what the hell Paul is trying to make with his desire to shim a couple tenths. You can't lift a V-way a thou without lifting the flat way off plane. Micro wedge maybe of non-flat and non contact in the flatway, but still there.

Scraping the headstock end of the ways alone will not fix his problem. Bitch TO scrape the headstock end of the ways to correct his believed error. Have to scrape a twist in the headstock end, and leave the rest of the ways "level". Or straight.

I have no problem with scraping where scraping is required. Some here think that when your lathe is cutting anything other than what you wanted, it needs scraped.

One in a hundred here can do this. Regardless you attended classes. Not criticizing the classes held, just that scraping is more than attending a couple weeks of beer and brats.

I have did this and I will tell you that scraping is a bitch of a job, and I would not wish it on any of you as a Summer Vacation.Scrape in a 24 foot crossrail on a VBM, or a 60 foot table way on a planer, and you will see that it is ALL sweatwork.

You got a 4 inch angle plate to practice on? BFD. Come back next year and tell us how you did on that 4 foot lathe bed. The one that you had to take out of service for a year while you scraped it in.

You need at least a 4 foot straightedge TO rub it in. You need, in some cases, 45 deg dovetail straightedge. You need calculus to determine how much you have to augment the Vee ways when you scrape them down to "true".

I am sure I will get grief from some who insist that an 18 inch straightedge is sufficient. I have had to scrape 40 foot long bedways with an 18 foot camelback that needed a crane to set it in place and 4 men to rub it in.

Don't believe everything you read in any forum. No Gods, here.

Cheers,

George

Forrest Addy
11-10-2009, 02:59 AM
George just blew in with the cold refreshing breeze of reality to this conflict of pet theories. I tell my scraping classes almost first thing that machine tool re-conditioning is a four year apprenticeship and the best that can be covered in a weekend class is an introduction to basic technique.

I read of people experencing for the first time the joys of chatter or part deflection or a chuck running out of true and the first thing they blame is the machine. They want to replace the spindle bearings, run Rollie's Dad's bag of tricks etc when the basic problem is up the sleeve of the operator or attempting to get silk purse performance from a saw's ear machine tool perhaps rescued from a chicken house. Or they will mount a handsome rugged new QC tool post and wonder why tolerance holding hasn't been magically transformed into a walk in the park.

This trade of ours is not easy, gentlemen. Mastering it even enough to support clock making, telescope building, support of car racing, home job shopping or whatever requires a major investment of time, diversified experience, and personal constructive humility. I served a four year apprenticeship rotating through the many sections of a large marine machine shop with diversions to outside machining, tool makers, instrument shop, etc. Was I anything special as a result? No. I was a basically competent but inexperienced journeyman. By the time I was pretty good 24 years later I was too old to really take advantage of it. Was I an all around machinist? No. I was a specialist as a machinist running large machine tools. The machinist trade is a house of many mansions meaning there are many specialties and advancement paths.

The milestones we are so proud of (first cut thread, first fitted male and female taper, first etc) are but baby steps worthy of congratulations perhaps but the next step is urgently beckoning and you have to move on to it. Complacency is the enemy and good enough is both the royal road to mediocrity and the efficient use of time and resources. Wisdom follows those who strike the most efficatious balance.

Accurate alignment of a lathe spindle is partly dependent on the bedways being in the same plane as when the machine was first manufactured and partly dependent on the machine's condition. The very LAST thing I would do is stick shims under the headstock when subsequent discussion reveals the machine is mounted in a trailer, a monument to elasticity subject to deflections caused by the operator merely shifting his weight.

I'm fond of saying "everything is made of rubber" as an illustration that no structure is perfectly rigid. Neither is a husky casting like a lathe bed subject to large deflections from small changes in loading. You have to keep a sense of proportion about these things.

Finally, qualitive adjectives such as those found in car magazines are the enemy of clear technical discourse. It may be boring and tedious but any discourse unaccompanied by numbers and a description sufficient to duplicate the problem for analysis renders understanding and technical problem solving almost impossible. How many original posters have casually added in the umpteenth message "BTW the lathe in located in a trailer" rendering pages of previous discussion not only moot but a waste of time.

There's a time and a place for everything including scraping and shims. The trick is to use the appropiate solution for a properly anaylized and vetted problem.

Since the lathe is in a trailer we have to go back to square one and reassess the design and stiffness of the trailer and how the lathe is mounted to it. Let's start over.

Paul Alciatore
11-10-2009, 03:28 AM
I may have missed this post, but how did the OP determine that the headstock was misaligned?

OK, OP here. How did I determint that I am cutting a taper? With a micrometer.

Seriously, I was cutting a 1/2" shaft on a saw arbor. The previously turned, one inch diameter head was gripped in a shop built collet. This collet has a shoulder that the head of the work piece rests against and I mounted the work with moderate pressure from the tailstock while tightening the collet. This should align the work in the collet. And yes, it was clean of chips and dings. The tailstock end was supported by a live center. The shaft was about 6" long and the tailstock end was small by about 0.005" or so. I call that a taper. No REASONABLE amount of offsetting of the tailstock would produce any correction so I concluded that the collet was holding it quite solidly - solidly enough to resist any influence of the tailstock - and the headstock is at a slight angle. If the part was deflecting while turning, it would be LARGE at the tailstock end, not small so I concluded it is not deflecting to any significant amount. Final cut was only a couple of thousanths so forces were minimal.

I do realize that the bed could be somewhat twisted. But could it be twisted enough to produce almost 0.001" taper per inch? I really doubt it is that twisted.

I plan to run some better tests with a piece of drill rod or shaft and a DI when I get the time. I should be able to detect the actual angle by taking several readings while rotating the headstock. I also want to re-check with the level as Evan suggests.

As for Forrest's suggestion of concrete pads under the trailer, that might be difficult. It is in a rental space and the landlord is a pain. I don't even want to think about asking, they would probably double my rent. In a year or two I will be moving anyway so it is really not worth the effort. When I get it in a permanent home, I will do a much better job of installing. For now, I am just trying to work with it as best as possible. If I find a serious twist, I may consider using a three point mount instead of the present four.

I DO appreciate all the discussion and suggestions. But I really just wanted a source for shim stock that comes in smaller graduations to experiment with. So, much thanks to all who have contributed.

macona
11-10-2009, 04:46 AM
Monarch used a test bar that fit into the taper in the spindle. Miller Machine sells test bars. Really needed to get a bed and head in alignment.

http://www.millermachineandfabrication.com/

Monarch 10EE beds are bedded to the base of the machine and are intended to never come apart (This is why you dont need to level a 10EE). Then the bed is ground (Monarch beds are flame hardened). The headstock is fitted to the bed using the test bar as a guide on how much must be remove from the V's on the bottom of the headstock. If I remember right both the large V and the small V continue under the headstock.





Lynn, imo, no way....because there is very little control over how much is removed where. you easily reduce the bearing area between the two parts




Very sound advice. shimming is not the way to do this job. they can move & squish but most importantly you do NOT get the level of control or granularity that you do with scraping. its quite easy to scrape a surface (light finish scraping) and not move the needle of a tenths indicator...in other words the depth of cut can be so fine that you have the control to nudge things a tenth one way or the other

....ok, but once stated, who cares if others dissent....you know some will waste a 100 pages of time on it, soooo moving on....

the question now is how to scrape this, I'd like to here some the experienced guys like Forrest explain how the axis of a high quality lathe like a monarch is scraped so that it will hold a tenth over some distance.

Is it as simply as getting bed spot on (levelled perfectly, no wear)....placing the headstock on and then with the carriage indicating a precise ground bar? Seems to me there has to be more to it; there's error in the bar, error in how the bar is held, etc. To get the HS axis to within a tenth over some distance would i think require resolution finer than a tenth.

...or do you put it all back together and turn a test piece, measure it with a indicating mic etc. that would overcome error in work holding, but limits the sample distance to being quite short. That sounds like a painful month of Sundays if it takes ten iterations!

So what are the options? how are the high quality lathes aligned?

Black_Moons
11-10-2009, 06:04 AM
yes it could easily be twisted enough to produce 0.001" per inch taper if NO attempt to align it has been made. I believe my lathe was that bad, the stand under the tailstock had to be moved several inchs back, and then the leveling feet adjusted maybe half a turn of the nut.

JCHannum
11-10-2009, 09:03 AM
After 87 posts, we are finally getting somewhere close to where we should have been at the outset. No lathe will be capable of accurate performance if not mounted properly. No amount of scraping or shimming will compensate for an unstable bench or flimsy floor.

If, as I recall, your lathe is a South Bend 9" or Light Ten, it is a bench mount and a sturdy enough bench can be fabricated that the flex of the trailer will not be a problem.

Put the lathe aside for the moment and either construct a new bench or reinforce the existing bench sufficiently to provide a rigid mounting surface. Construction grade lumber, bolted together, (I would suggest 2x4 or 2x6 for the top, with similar stringers and 4x4 legs at a minimum) or structural steel angle and plate welded up. There are many examples floating around for inspiration. If using a wood bench top, I recommend steel flats under the legs to prevent their sinking into the wood over time.

Once that is completed, the process of setting up and aligning the lathe can begin. Don't take shortcuts or expedients here either. Use the best level available and use it properly on the ground surfaces at the headstock and tail end of the lathe. Any other method will introduce other errors into the setup.

Unless the cause of the problem is corrected any other attempted fixes will only result in other problems and you will end up constantly chasing your tail, piling patch upon patch. At some point, you will have to address the underlying cause. It is best to do it at the outset.

pcarpenter
11-10-2009, 10:40 AM
I have to agree that my experience follows JC's.....trying to fix a "problem" at anything but the real source seems to inflict yet another problem somewhere it should not. Lane recommended several pages back that the OP should look at twist. For a small lathe, in a trailer, a "concrete base in a box" could be poured to give some rigidity. Frame it in 2x construction lumber and pour yourself a brick to set the lathe and stand on and then make the bed planar using a level and adjustment screws on its base or table.

There are lots of ways to end up with taper. That doesn't mean that you can pick the one that's most convenient to address, in an effort to fix the problem. For example, shimming a headstock to a twisted bed doesn't fix the problem properly like untwisting the bed. You will make the headstock/bed/tailstock assembly turn without taper in only the configuration you use for testing. Twist, by definition, does not occur at only one point. It means that the bed is misaligned to the headstock in an inherently variable way along it's length.

I recall a guy trying to sell me a big LeBlond lathe with worn tailstock ways. These typically wear in a ramp fashion toward the headstock. He insisted that he had fixed it by shimming the tailstock...something he had to do to turn big tractor pins. I reminded him that this only made that tailstock correct at the point he chose to shim it.

I agree with JC on another issue:


I highly recommend the Connelly book to anyone with a desire for information on the proper methods of inspecting machine alignment and the sequence these should be done in. However, simply reading this or any book does not an expert make.



You don't *read* "Machine Tool Reconditioning"...you consult with it:D There is years of information in there...but it's not even a good "reference" book in that you don't go look up the concavity spec for a lathe facing cut in a table. That sort of information is in there, but so much of it is in with all sorts of other minutia...good minutia mind you...but it's more like Connelly did a brain dump.

Paul

Evan
11-10-2009, 11:12 AM
No lathe will be capable of accurate performance if not mounted properly. No amount of scraping or shimming will compensate for an unstable bench or flimsy floor.


Finally, something we can agree on. It isn't that hard to make a solid and very rigid mount for a light lathe. Note the piece of 10x2 channel iron that the lathe is bolted to. The bench top under it is a 2x4 triangulated frame with 3/4" plywood glued and screwed to both sides. The rest of the bench is 2x4 frame with 3/4" high density particle board screwed to the sides and back and a reinforced shelf to square it up.

http://metalshopborealis.ca/pics/shop1a.jpg

Paul Alciatore
11-10-2009, 11:44 AM
AAAAAaaaarrrrrrrrrgh!

OP here again.

I totally agree that a lathe should be properly mounted. Neither my welded steel bench nor something like Evan's steel channel base and carefully constructed table will eliminate this requirement. Frankly, the CI bed is probably less flexable than either of these just because it is CI. Personally I wouldn't trust any wood construction as wood can change dimensions with humidity levels. One piece, solid concrete slab and steel table is the only way to go IMHO. I don't even like welded construction as it can contain stresses that will relax over time. Anybody got an oven big enough to put a whole table in? Come on, anybody got one? And any table should be tied down, in ONE place on that concrete pad. If it moves an inch, it will twist and change as concrete is not perfectly flat.

But, due to my landlord situation, I can not do that now. I would if I could. REALLY, I would. But I can't.

Hence, I need some kind of expedient solution that is not optimal, but that would help, at lease somewhat for the time being. I am not stupid enough to scrape or otherwise remove metal on the lathe under these circumstances as that would do more damage than good and would just be hard to reverse later when the mount is right.

Therefore, anybody got any better solutions than shimming? Or, WHERE CAN I GET FINER SHIMS?

Richard Wilson
11-10-2009, 12:15 PM
I did exactly that to my SB9 Paul to reduce the tendency to turn concave on a facing cut. It needed about .001" brass shim stock on the back Vee way to bring it into line. It made so little difference to the height that there was no reason the alter the tailstock. If you take the .001 x .707 = .0007 /2 = .00035 that is how much it raised the centre line. Raising the centre line by .0003 has no measurable effect on anything.

BTW, Glad plastic Cling Wrap is .0005 thick and is pure high density polyethelene.

I understood that when lathes were new that they did face slightly concave, in middle age they wore to facing flat, and only when geriatric would they face convex. I think the normal limits for lnew athes shows this as acceptable. Bolting 2 slightly concave surfaces together isn't usually a problem.

Richard

pcarpenter
11-10-2009, 12:17 PM
Shim away, but even if you can't make a more rigid mount for the lathe you still need to try to make it's bed ways as flat as possible before you start into propping up the headstock. Do as much of the shimming as you can under the "feet" and not between the headstock and bed.

As I pointed out, your lathe bed likely looks like a piece of wet lasagna if not levelled properly. What point on that twisted piece are you going to use as the reference for your headstock? That's the point near which you will get what you are after....anywhere else is anyone's guess.

Edit-- by the way, Evan's design has a lot more rigidity with the lathe shimmed for squareness and fastened to the channel. You would be amazed at cast iron's ability to flex under it's own weight. A short, light lathe is less problematic due to it's rigidity vs. weight trying to pull it out of alignment, but they all twist....sometimes many thousandths. The "rigid" cast iron ram on my Bridgeport mill flexed 6 *thousandths* when I bolted the head on it, creating no ends of grief for me since I had scraped the knee without the head in place. My floor lathe started with so much twist at the beginning of the levelling process that I could not even use a precision level at the start. It was off by many thousandths from one end to the other and the bubble never moved. Trying to fix a twist down toward the tailstock end by twisting the head to match, means variable amounts of misalignment all along the bed.

That having been said, I recall that the mylar baloon material is pretty thin....the stuff they make those silvered "printed message ballons out of.
Paul

Evan
11-10-2009, 12:49 PM
The bench is so rigid that it can be propped up on one foot so it rocks and there is no measurable change in the squareness of the lathe. Also, the humidity in the basement is always within a ten percent range of about 30 to 40%. The lathe is securely bolted to the channel iron but the channel iron is only lightly fastened to the bench to prevent it moving about. The lathe and the channel iron form a bridge that is orders of magnitude stiffer than the lathe alone.

The concrete of the floor doesn't move either. I discovered this summer that I have the hardest concrete in the world. The aggregate is entirely small corundum stones. Under that is 100 feet of hardpan resting on megalithic rock. I know that the floor doesn't move because my seismograph is an extraordinarily sensitive tilt meter. It can record the tilt caused by a person walking near the house. It only changes it's balance when we have a weight change such as a rain or snowfall.

Mcgyver
11-10-2009, 04:35 PM
The bench is so rigid that it can be propped up on one foot so it rocks and there is no measurable change in the squareness of the lathe.

how is this determined, by making test cuts while the lathe is balanced on your foot?

Evan
11-10-2009, 04:41 PM
See the climbing piton on the floor at the lower left? It is shimming (!!) the corner of the table so it won't rock. Without it or with it the only difference is whether it rocks or not. A stressed skin panel like the benchtop is an incredibly strong structure. It's like having a piece of plywood 5 inches thick.

oldtiffie
11-10-2009, 05:13 PM
Once again we are getting away from the requirements of the OP and his topic and telling what we would do on our lathe in our shop.

Paul.

Isolate the lathe from the trailer/container it is in by removing all bolts fastening it to the trailer/container so that the lathe is "free" of any external forces or constraints.

Have the lathe supported on no more than three points (like a tri-pod). It may mean jacking the lathe up so that the head-stock end is supported on two jacks and the two legs at the tail-stock end "bridged" and supported on the bridge with the bridge supported on a single central jack.

Make the best of what you have until the lathe is in/on a more permanent and solid (read: heavy, concrete slab etc.) base/foundation structure.

Its a classic case of making do with what you have as an interim measure.

John Stevenson
11-10-2009, 05:50 PM
The bench is so rigid that it can be propped up on one foot so it rocks and there is no measurable change in the squareness of the lathe. Also, the humidity in the basement is always within a ten percent range of about 30 to 40%. The lathe is securely bolted to the channel iron but the channel iron is only lightly fastened to the bench to prevent it moving about. The lathe and the channel iron form a bridge that is orders of magnitude stiffer than the lathe alone.

The concrete of the floor doesn't move either. I discovered this summer that I have the hardest concrete in the world. The aggregate is entirely small corundum stones. Under that is 100 feet of hardpan resting on megalithic rock. I know that the floor doesn't move because my seismograph is an extraordinarily sensitive tilt meter. It can record the tilt caused by a person walking near the house. It only changes it's balance when we have a weight change such as a rain or snowfall.


It must be most reassuring knowing the floor is so level and secure when you are wading thru all that bullsh*t. :rolleyes:

Is that 100 feet of hardpan accurate to 3 or 4 microns.

.

tattoomike68
11-10-2009, 05:57 PM
Shimming the headstock to match bed wear is nuts, its all the worn stuff on the ass-end that have wear. (you can knock a head stock off if you thread into it with a 2 tpi at 2,000 rpm)

myself I ran real crap lathes at some jobs but still made good parts, I use a file and some worn emory cloth to hit less than .0005"


you can chase .0001 forever and never do anything but burn time.

I want to know what makes you think its the headstock? Did you crash the lathe? if that were the case you might see .010" taper per inch.

Mcgyver
11-10-2009, 05:59 PM
The concrete of the floor doesn't move either. I discovered this summer that I have the hardest concrete in the world. The aggregate is entirely small corundum stones. Under that is 100 feet of hardpan resting on megalithic rock. I know that the floor doesn't move because my seismograph is an extraordinarily sensitive tilt meter. It can record the tilt caused by a person walking near the house. It only changes it's balance when we have a weight change such as a rain or snowfall.

...... short of sampling all other concrete how did you determine yours was the hardest in the world?

oldtiffie
11-10-2009, 06:29 PM
Lottsa "Viagra" as a hardener.

Evan
11-10-2009, 06:34 PM
It must be most reassuring knowing the floor is so level and secure when you are wading thru all that bullsh*t.

Is that 100 feet of hardpan accurate to 3 or 4 microns.

.


About a foot or so. It's recorded in the well drilling log. The well is 350 feet deep.
[edit] You really must learn to avoid making such an ass of yourself John. I imagine it's too late though.


short of sampling all other concrete how did you determine yours was the hardest in the world?

The only mineral harder than corundum is diamond.

tattoomike68
11-10-2009, 06:51 PM
Lottsa "Viagra" as a hardener.


Viarga is scarry, what do you do with an erection that last 4 hours? Dress a grinding wheel with it?

LOL :D

Mcgyver
11-10-2009, 07:01 PM
The only mineral harder than corundum is diamond.

so? a big logical stretch between that truth and your's is the hardest concrete in the world, no?

Evan
11-10-2009, 07:09 PM
No stretch. It took three weeks working every day with a 2 inch stroke air hammer to pound a hole through the footing for a drain pipe. It should have taken a day or two. I have drilled concrete before. Why do you care?

S_J_H
11-10-2009, 07:22 PM
Paul, does your SB9 have the 2 piece stand at the tailstock end? If you do , there are setscrews on either side that you can use to level out the bed as desired.
My SB9 has this feature and while using a level on the cross slide to traverse the bed it is quick and easy to get the bed level on our little machines.

Steve

oldtiffie
11-10-2009, 07:23 PM
Originally Posted by oldtiffie
Lottsa "Viagra" as a hardener.


Viarga is scarry, what do you do with an erection that last 4 hours? Dress a grinding wheel with it?

LOL :D

1.
Go and show it off - or find a use for it:
http://i200.photobucket.com/albums/aa294/oldtiffie/Funnies/Strolling1.gif

2.
If not day-light - turn the switches on - all of 'em!!
http://i200.photobucket.com/albums/aa294/oldtiffie/Funnies/Viagraswitch1.jpg

QED

I'd be interested to see how Evan took the pics though!!

lazlo
11-10-2009, 07:24 PM
In Texas we use diamond aggregate in our concrete. So we can measure far better than 3 - 4 microns.

tattoomike68
11-10-2009, 07:40 PM
Concrete is never hard to drill unless you used 304 stainless rebar and old safes as rebar. drilling concrete is always gravy work.

S_J_H
11-10-2009, 07:41 PM
This is what I was referring to-
http://i109.photobucket.com/albums/n48/S_J_H/SOUTHBEND%209/IMG_1663.jpg

wierdscience
11-10-2009, 08:01 PM
Paul,did you try it with a different center? Live centers will offset in their own bearings without any help from tool pressure.

lane
11-10-2009, 08:04 PM
Even a leveled lathe will cut a taper . If bed is twisted. and a south bend bed will twist easy. My 10 k would never cut straight out of the box. bought it brand new. Get it level best you can. Then adjust the tailstock end to make lathe cut straight. Their is a adjusting mechanize their to do it. The level reads off two lines but lathe turns straight with in .0002 in 8 inches turning and adjusting the two collar method. Any more than that you use the tail stock and adjust it to turn straight.
Even my 13x40 ACCER is not dead level but turn`s with in .0001 in 12 inches With two collar test and 2 inch diameter shaft fitted to spindle taper .

Evan
11-10-2009, 08:06 PM
Concrete is never hard to drill unless you used 304 stainless rebar and old safes as rebar. drilling concrete is always gravy work.


Unless it has corundum as aggregate. I sharpened up a piece on my diamond wheel and it will scratch tungsten carbide. It is also incredibly impact resistant. I could take a rock and put the chisel directly on it and give it a burst. Instead of breaking the rock it blunted the chisel and the chisels are US made. I could take the same air hammer and put it against a 4 inch slab from the walkway that I pulled out and in a single burst knock off a big chunk the size of your foot. Trying to drill through the footing turned into an exercise in powdering the cement around the aggregate until the aggregate pieces would fall out.

John Stevenson
11-10-2009, 08:10 PM
[edit] You really must learn to avoid making such an ass of yourself John. I imagine it's too late though.



Reminds me of the story when this guy on a pier is fed up of this guy bragging he's got the best, the biggest, the most expensive and just then the QE2 sails past.

First guy rushes to the end of the peir and shouts "Come in number 4 your times up "

As regards asses, that's another one that's whooshed over you head unnoticed to you but not a lot of other readers.
:D

You really need to get out more and hone your social skills, talking to that dawg isn't helping.

.

lane
11-10-2009, 08:30 PM
Quote South Bend (Turning Test For Leveling
A simple turning test can be used to check the leveling of the lathe . To make the test a bar of steel one inch or larger in diameter in the chuck and machine two collars of equal diameter three are four inches apart. then take a cut not to exceeding .002"in depth across both collars' using a fine feed that will not crowd the bar from the tool. Measure both collars if diameter is equal lathe IS level. A difference in diameter indicates lathe is not level . Do this
If the diameter of the outer collar is grater than that of the inner collar ,shim under the front of the lathe at the tailstock end until both collars are turned the same diameter. If the diameter of the inner collar is greater than the outer collar ,shim under the back of the lathe at the tailstock end until both collars are turned the same diameter . Set level on end of lathe and move 1/2 line at a time and take cut . 2 or 3 cuts and you should have it . If you have the adjusting screws SJH talked and showed use them with level

Mcgyver
11-10-2009, 08:31 PM
Unless it has corundum as aggregate. I sharpened up a piece on my diamond wheel and it will scratch tungsten carbide.

true self sufficiency, not only can you make tools bits out of your floor, you can make grinding wheels for your carbide as well, amazing :D

Evan
11-10-2009, 08:39 PM
Don't laugh too much. I shaped one piece into a tool bit and it worked on a grade 8 bolt. It wouldn't hold an edge though. :cool:

J Tiers
11-10-2009, 09:20 PM
OK, OP here. How did I determint that I am cutting a taper? With a micrometer.

Seriously, I was cutting a 1/2" shaft on a saw arbor. The previously turned, one inch diameter head was gripped in a shop built collet. This collet has a shoulder that the head of the work piece rests against and I mounted the work with moderate pressure from the tailstock while tightening the collet. This should align the work in the collet. And yes, it was clean of chips and dings. The tailstock end was supported by a live center. The shaft was about 6" long and the tailstock end was small by about 0.005" or so. I call that a taper. No REASONABLE amount of offsetting of the tailstock would produce any correction so I concluded that the collet was holding it quite solidly - solidly enough to resist any influence of the tailstock - and the headstock is at a slight angle. If the part was deflecting while turning, it would be LARGE at the tailstock end, not small so I concluded it is not deflecting to any significant amount. Final cut was only a couple of thousanths so forces were minimal.

I do realize that the bed could be somewhat twisted. But could it be twisted enough to produce almost 0.001" taper per inch? I really doubt it is that twisted.

I plan to run some better tests with a piece of drill rod or shaft and a DI when I get the time. I should be able to detect the actual angle by taking several readings while rotating the headstock. I also want to re-check with the level as Evan suggests.

As for Forrest's suggestion of concrete pads under the trailer, that might be difficult. It is in a rental space and the landlord is a pain. I don't even want to think about asking, they would probably double my rent. In a year or two I will be moving anyway so it is really not worth the effort. When I get it in a permanent home, I will do a much better job of installing. For now, I am just trying to work with it as best as possible. If I find a serious twist, I may consider using a three point mount instead of the present four.

I DO appreciate all the discussion and suggestions. But I really just wanted a source for shim stock that comes in smaller graduations to experiment with. So, much thanks to all who have contributed.

Well you may have just answered all your questions (except the shim stock).

but I have a question or two.....

1) when you were turning that saw arbor, was the initial cut obviously interrupted or off-center? If the part was rotating off-center, it should have been.

Assuming the answer to #1 is "no", then you have, for some reason, the axis of the spindle pointing elsewhere, other than at a properly centered tailstock and at an angle to the true travel of the carriage.

This COULD be for several reasons, including:

A) bed is twisted. Since there is exactly ONE way for it to be perfect, and a very large number of ways it can be messed up, it almost certainly IS twisted, and you may want to assess how MUCH it is twisted. There is no law that says that is the only problem you have. The twist may not be enough to worry about, or it may be quite large. No measurement, no data......... no data, no can fixee.

B) headstock has crap under it holding it out of line. You've had it apart, apparently, and I will trust your intelligence and assume this is not true...... (In MY case that mightn't be so wise, I left a piece there once, and had to clear it out later)

C) headstock is for some unknown reason actually out of line. I suppose this is possible, but it is starting to fall into the "bent spindle game" category of 'find the most expensive problem and assume that is it"....

D) lathe was made wrong..... even farther into that forbidden territory

My money is on the bed twist. Unless you KNOW it isn't twisted, it is far better to assume it IS, because.........." there is only one way for it not to be".........

Can it cause 0.001"/inch? Sure, why not? it only has to move the centerline 5 tenths per inch to do that. Do you know it is NOT twisted to that degree of accuracy?

WHAT TO DO?

1) possibly live with it...... but if you need it to cut right, you probably want to fix it.

if you live with it, you can always fix the tapers by offsetting the tailstock AND USING CENTERS, not the chuck, collets, etc.

2) Make a stand that takes the nasty floor out of the picture. I don't want to get into the corundum and whatever wars... . But surely you can make a solid table that your lathe can sit on, and be shimmed (under the feet) into straightness on. A "butcher block table" made of 2 x 4s is possible, as is laminated plywood, a piece of 1" steel, concrete, or whatever your inclination is. If you use wood, DO use the metal plate plan shown in an earlier pic..... works quite well

it isn't as if that will be of no use after you move..... it surely will.

3) Shim the headstock or whatever to get the taper out.

I think that is the hard way, the long way, and the one that will cause you the most frustration when you need to change the shims every so often..... but you CAN do that.....

darryl
11-10-2009, 09:37 PM
Ok, Paul I have a suggestion. I don't know what your stand is like, but I'll assume it has four legs. I'm going to turn it into three.

On the tailstock end, put a brace between the legs quite near the floor. Make it a solid piece since you will be adding a foot to the center of that that will carry the entire weight of that end. In use you will adjust that foot to just barely lift off the other two feet.

Add a brace from each leg under the headstock end that will go to the center point of the first brace, preferably directly to the center legs axis. One or both of these braces will be made adjustable for length- maybe a turnbuckle can be used, but you will also need to be able to run a lock nut on both sides of the turnbuckle.

Have the lathe mounted as you normally would to the bench.

Two things you need to check now, one is for a sag or an upward bow across the length of the bed. By adjusting both turnbuckles you can bow the legs in or out and effect how the flatness of the top is that the lathe is bolted to, and hence the curve of the lathe bed. The other thing to check is for parallellism of the ways. That can be adjusted by lengthening one turnbuckle and shortening the other by the same amount. Forget about shimming the mounting feet of the lathe, just bolt it securely to the bench before beginning any of this. The only shimming should be if the lathe and the benchtop are obviously out of whack with each other. You don't want to stress one or the other just to get a good contact patch between them. The mounting bolts should just be there to keep the two pieces together, not warp one to fit the other.

Now raise the tailstock end by running that single foot down to lift both feet off by a half inch or so- this gives you room to put some kind of soft rubber pads under those two legs. Lower the single leg now to put only a little down pressure on those pads. If the legs don't both touch the floor at the same time, use some shims to get it that way, or maybe have adjustable feet there. Finally, raise the headstock end and put some spacers under both feet to cater to the level of the lathe in the traditional carpenter sense. No rubber pads needed there unless you want to, but not thick. This would be more to keep the feet from skidding around on the floor.

Some of my reasoning- the two soft pads under the tailstock end legs are only there to keep the lathe from falling over if it gets pushed or forced in some way. A single leg tips too easily, whereas this way you have the two legs that can stabilize, but only do so when needed. The triangulation keeps any twisting of the trailer floor from transferring a twist to the stand and hence to the lathe. Not having to play around with shimming the feet of the lathe to the benchtop lets you secure it solidly so the lathe and the bench become more or less one structure. I will concede that warping of the bench due to moisture changes in the wood may still be a factor in a shifting alignment.

One thing that will be important for this idea to work is for the tail end legs to not be braced any other way than by the triangulation method I've outlined, except to each other. No braces that run from left to right.

What we've done here is allowed the trailer floor to lean the lathe front to back, and/or left to right, but not to twist it. I don't see a problem to secure the head end feet to the floor so the lathe doesn't move around, but you can't prevent the single foot at the tail end from sliding- you have to let it slip if it needs to. Lower the feet at that end only enough to get some contact of the rubber pads, but not enough that they exert any control over the stand. That's the whole thing you're trying to prevent. They should only come into play to prevent an accidental tipover.

Obviously, much depends on how the stand is built.

Mcgyver
11-10-2009, 10:06 PM
Paul, as partial recompense for all the silliness, I just tried some gum wrappers and they're about .0002" using my best mic.

it'll be tough to work with, it'll crinkle in a heartbeat then you're .0004. its difficult enough to handle without creasing or tearing that, i dunno, I'd probably figure something else out or look for stuff at least a thou but with .0001 increments. or scrape - you could get a chunck of Evans floor if you don't have a scraper blade :)

Brand of gum was cobalt 5 iirc. Innovative packaging, in so far as gum packaging can innovative. black cardboard box with a flap, blue foil wrappers. burn the paper off and there you go.

firbikrhd1
11-10-2009, 11:03 PM
While I've enjoyed all of the various opinions here about shimming v.s. leveling, etc. and wholly agree about a stable base for a lathe, there's no way I'm getting into the foray regarding which is the best thing to do.

With regard to thin shims, this may or may not meet your requirements: Contact a Snap-On Tool and ask him to get you 12" feeler gauges. They are sold individually and can be had in .0015 as the thinest size. Mac Tools may have them as well as might Matco. They may even be available in thinner sizes but I can't say for certain. Cost won't be anything like obtaining a shim stock assortment.

Nestle's Quick has a thin, silver Mylar like seal on their plastic powdered chocolate containers that may be of use as well but I've never measured the thickness.

Hope this helps, on with the discussion.....

Mike Burdick
11-11-2009, 12:22 AM
My question is just that. How would you shim something a few tenths? ...
Paul,

I have never used anything like this so I don't know how thick copper leaf is. But ....

http://www.lagoldleaf.com/images/xsect1/cpr_img_014a.jpg

It's not expensive if one Book would be enough ($5.00). I don't know if the thickness is consistent but that probably wouldn't matter and perhaps work to your advantage - kind of a variety pack.

Paul, regardless of what material you use, please let us know if your experiment works out. After all, by using shims, one will not do irreparable damage to the machine.

http://www.lagoldleaf.com/product_info.php?cPath=28&products_id=111

Paul Alciatore
11-11-2009, 02:08 AM
Once again we are getting away from the requirements of the OP and his topic and telling what we would do on our lathe in our shop.

Paul.

Isolate the lathe from the trailer/container it is in by removing all bolts fastening it to the trailer/container so that the lathe is "free" of any external forces or constraints.

Have the lathe supported on no more than three points (like a tri-pod). It may mean jacking the lathe up so that the head-stock end is supported on two jacks and the two legs at the tail-stock end "bridged" and supported on the bridge with the bridge supported on a single central jack.

Make the best of what you have until the lathe is in/on a more permanent and solid (read: heavy, concrete slab etc.) base/foundation structure.

Its a classic case of making do with what you have as an interim measure.

OK, that is a suggestion that I can accomplish. I had something like that in mind when I said a three point suspension. At present, the headstock end it torqued down and the bolts at the tailstock are somewhat loose. As I said, I will try the level again just to see where it is.

gmatov
11-11-2009, 02:12 AM
Paul,

Build a new table. Maybe 6 or 8 inch pipe, 3 or 4 feet, fab or steel plate for the top.

Pedestal. ONE point to the floor, and I don't care if, after you align the thing, you lay the trailer on it's side, the lathe will be as true as you had it when it was horizontal. Just as it would be if it was on a destroyer rolling 60 degrees in a storm with need for a new whatever "right NOW!!!"

Tailstock end, make a third point for a three point mount. HS bolted down, tailstock end on a single point, IF the lathe can normalize to the "as made" condition. Tee shaped. 2 bolts to anchor to the Tee, one bolt to anchor to the table.

3 legs will always be more stable than 4.

Boats twist in normal sailing, they twist a hell of a lot more in heavy weather. My wooden boat, Minesweeper, made lots of noise when we came up from the Crib in heavy weather. Tropical storm weather.

Machinists could hold size if they could hold their lunch down, because the lathe mount was not 4 points attached to the ship's ribs.

Once upon a time, in the Westinghouse, rotor slotter operator was fired for screwing up a slot in a 100 tons of rotor. Investigation proved that the OEC with another 100 ton load actually twisted the bed of the 200 ton machine when the EOC passed over the slotter. Machine was about 125 foot long. Twisted like Forrest's "pretzel".

Couple thou qualified as "pretzel". Scrap.

Gotta be more to why Paul's machine is so off, and I don't know if it is the mounting. Loose the tailstock bolts and see if it tries to go back to "straight after a few days.

Your level does not have to read level, it has to read the same at both ends.
Cheers,

George

Paul Alciatore
11-11-2009, 02:13 AM
Viarga is scarry, what do you do with an erection that last 4 hours? Dress a grinding wheel with it?

LOL :D

Would you care to demonstrate that technique?

Paul Alciatore
11-11-2009, 02:15 AM
Paul, does your SB9 have the 2 piece stand at the tailstock end? If you do , there are setscrews on either side that you can use to level out the bed as desired.
My SB9 has this feature and while using a level on the cross slide to traverse the bed it is quick and easy to get the bed level on our little machines.

Steve

No, sorry, just a one piece stand.

Paul Alciatore
11-11-2009, 02:28 AM
Concrete is never hard to drill unless you used 304 stainless rebar and old safes as rebar. drilling concrete is always gravy work.

Oh yea? I beg to differ. When I was in high school, I watched a work crew try, that's TRY to drill holes in the 70 or 80 year old concrete floors for some sprinkler pipes they had to install. They started on the second floor one day and when I came down for lunch they were still in the same hole. Ditto when I went home. Then day two, same thing, SAME HOLE, no penetration. And yes, they used power tools. They didn't break that FIRST hole through to the first floor until mid-day on the third day. The finished hole ran through about two feet of concrete. I don't know the details of the kind of drill, but I do know that that job took quite a while. Five story building, city block long and "U" shaped and the installation of the sprinklers ran all over the building. They were there quite a while.

Good concrete. Well cured concrete.

I should rent a room there for my shop.

Paul Alciatore
11-11-2009, 02:49 AM
Well you may have just answered all your questions (except the shim stock).

but I have a question or two.....

1) when you were turning that saw arbor, was the initial cut obviously interrupted or off-center? If the part was rotating off-center, it should have been.

Assuming the answer to #1 is "no", then you have, for some reason, the axis of the spindle pointing elsewhere, other than at a properly centered tailstock and at an angle to the true travel of the carriage.

This COULD be for several reasons, including:

A) bed is twisted. Since there is exactly ONE way for it to be perfect, and a very large number of ways it can be messed up, it almost certainly IS twisted, and you may want to assess how MUCH it is twisted. There is no law that says that is the only problem you have. The twist may not be enough to worry about, or it may be quite large. No measurement, no data......... no data, no can fixee.

B) headstock has crap under it holding it out of line. You've had it apart, apparently, and I will trust your intelligence and assume this is not true...... (In MY case that mightn't be so wise, I left a piece there once, and had to clear it out later)

C) headstock is for some unknown reason actually out of line. I suppose this is possible, but it is starting to fall into the "bent spindle game" category of 'find the most expensive problem and assume that is it"....

D) lathe was made wrong..... even farther into that forbidden territory

My money is on the bed twist. Unless you KNOW it isn't twisted, it is far better to assume it IS, because.........." there is only one way for it not to be".........

Can it cause 0.001"/inch? Sure, why not? it only has to move the centerline 5 tenths per inch to do that. Do you know it is NOT twisted to that degree of accuracy?

WHAT TO DO?

1) possibly live with it...... but if you need it to cut right, you probably want to fix it.

if you live with it, you can always fix the tapers by offsetting the tailstock AND USING CENTERS, not the chuck, collets, etc.

2) Make a stand that takes the nasty floor out of the picture. I don't want to get into the corundum and whatever wars... . But surely you can make a solid table that your lathe can sit on, and be shimmed (under the feet) into straightness on. A "butcher block table" made of 2 x 4s is possible, as is laminated plywood, a piece of 1" steel, concrete, or whatever your inclination is. If you use wood, DO use the metal plate plan shown in an earlier pic..... works quite well

it isn't as if that will be of no use after you move..... it surely will.

3) Shim the headstock or whatever to get the taper out.

I think that is the hard way, the long way, and the one that will cause you the most frustration when you need to change the shims every so often..... but you CAN do that.....


The answer to #1 is no. The initial cut was pretty much centered, not interrupted. This is exactly why I am concluding that the headstock is not aligned to the bed. Or the bed is twisted.

1B. I took great care when originally installing to have all surfaces clean and I found nothing when I recently removed it. Again, I took great care for all to be clean and I got much the same results when reassembled.

1C. This is my great suspicion. I did buy the lathe used. It appears to have been assembled from various pieces and that would mean that the headstock was not properly scraped to the bed.

1D. Same answer as 1C.

2. Wouldn't a three point suspension take the table out of the equation?

3. Shimming. I'm not sure why it would be the most difficult. It may not be satisfactory, I know that. But it seems quite simple. And easily reversable. And I am already frustrated.

Paul Alciatore
11-11-2009, 02:55 AM
Ok, Paul I have a suggestion. I don't know what your stand is like, but I'll assume it has four legs. I'm going to turn it into three.

On the tailstock end, put a brace between the legs quite near the floor. Make it a solid piece since you will be adding a foot to the center of that that will carry the entire weight of that end. In use you will adjust that foot to just barely lift off the other two feet.

Add a brace from each leg under the headstock end that will go to the center point of the first brace, preferably directly to the center legs axis. One or both of these braces will be made adjustable for length- maybe a turnbuckle can be used, but you will also need to be able to run a lock nut on both sides of the turnbuckle.

Have the lathe mounted as you normally would to the bench.

Two things you need to check now, one is for a sag or an upward bow across the length of the bed. By adjusting both turnbuckles you can bow the legs in or out and effect how the flatness of the top is that the lathe is bolted to, and hence the curve of the lathe bed. The other thing to check is for parallellism of the ways. That can be adjusted by lengthening one turnbuckle and shortening the other by the same amount. Forget about shimming the mounting feet of the lathe, just bolt it securely to the bench before beginning any of this. The only shimming should be if the lathe and the benchtop are obviously out of whack with each other. You don't want to stress one or the other just to get a good contact patch between them. The mounting bolts should just be there to keep the two pieces together, not warp one to fit the other.

Now raise the tailstock end by running that single foot down to lift both feet off by a half inch or so- this gives you room to put some kind of soft rubber pads under those two legs. Lower the single leg now to put only a little down pressure on those pads. If the legs don't both touch the floor at the same time, use some shims to get it that way, or maybe have adjustable feet there. Finally, raise the headstock end and put some spacers under both feet to cater to the level of the lathe in the traditional carpenter sense. No rubber pads needed there unless you want to, but not thick. This would be more to keep the feet from skidding around on the floor.

Some of my reasoning- the two soft pads under the tailstock end legs are only there to keep the lathe from falling over if it gets pushed or forced in some way. A single leg tips too easily, whereas this way you have the two legs that can stabilize, but only do so when needed. The triangulation keeps any twisting of the trailer floor from transferring a twist to the stand and hence to the lathe. Not having to play around with shimming the feet of the lathe to the benchtop lets you secure it solidly so the lathe and the bench become more or less one structure. I will concede that warping of the bench due to moisture changes in the wood may still be a factor in a shifting alignment.

One thing that will be important for this idea to work is for the tail end legs to not be braced any other way than by the triangulation method I've outlined, except to each other. No braces that run from left to right.

What we've done here is allowed the trailer floor to lean the lathe front to back, and/or left to right, but not to twist it. I don't see a problem to secure the head end feet to the floor so the lathe doesn't move around, but you can't prevent the single foot at the tail end from sliding- you have to let it slip if it needs to. Lower the feet at that end only enough to get some contact of the rubber pads, but not enough that they exert any control over the stand. That's the whole thing you're trying to prevent. They should only come into play to prevent an accidental tipover.

Obviously, much depends on how the stand is built.

This sounds like it just might work. I need to buy some steel.

J Tiers
11-11-2009, 03:26 PM
The answer to #1 is no. The initial cut was pretty much centered, not interrupted. This is exactly why I am concluding that the headstock is not aligned to the bed. Or the bed is twisted.

1B. I took great care when originally installing to have all surfaces clean and I found nothing when I recently removed it. Again, I took great care for all to be clean and I got much the same results when reassembled.

1C. This is my great suspicion. I did buy the lathe used. It appears to have been assembled from various pieces and that would mean that the headstock was not properly scraped to the bed.

1D. Same answer as 1C.

2. Wouldn't a three point suspension take the table out of the equation?

3. Shimming. I'm not sure why it would be the most difficult. It may not be satisfactory, I know that. But it seems quite simple. And easily reversable. And I am already frustrated.


Eh... I mean make a good solid table top, which you can still use after you move. Then 2 points, 3 points, 4 points, all will be fairly irrelevant. the feet the whole thing sits on will be much less of an issue than getting the lathe straight where it sits on the table. The combination of bed and table can be made to be quite rigid.

As for converting the lathe to 3 point, well, OK..... But usually that works best with a very substantial rigid bed. S-B smaller machines, which I understand you have, are not very rigid, and might cause troubles later, probably chatter, etc. and it assumes that the bed is actually straight to start with, which may or may not be true. If not, then with 3 points you'll never force it back. (I know, that's supposed to be bad, but what's the alternative?)

First things first.... check the bed for twist with a decent level of no worse than 0.005/foot sensitivity (a really good one may just "peg" and be of no help). Then you'll have a chance. We're all guessing until that is done.

Evan
11-11-2009, 04:05 PM
The SB9 is nowhere close to being rigid enough to use a three point support for the lathe itself. If you bolt down the head end and leave the tail free you can twist it visibly with your hands. You can even shove or pull it to the back or front and parallelogram the bedway. That is why I bolted mine to the channel iron and then shimmed the feet to bring the bed planar and square.

koda2
11-11-2009, 09:16 PM
Don't know if this was mentioned but here is some stainless shim that was recommended to me.
http://www.hasberg-schneider.de/english/techinfo.html

Or go here:
http://www.home-machine-shop.com/products/products-hasberg-shim-stock.htm

Minimum is .00019

Dave A.