PDA

View Full Version : What Do You Do When The OLD GIRL STARTS TO SAG



JoeLee
06-21-2010, 11:14 PM
After responding to a recent post where someone asked why the lathe cut a slight taper on one end of a short piece of work, I started thinking why my lathe is showing signs of the same thing.
I always thought it was because the bed ways are worn more right in front of the chuck where most of your work is done. However after checking the bed with a level I noticed that the center of the bed has started to sag over the years. In the past I have heard a few Clausing 5900 owners comment that "they made a good machine but they should have beefed up the bed a little more" Now I see why.
I don't think there is really much I can do about it other than drill a hole in the center of the chip pan and post the center......... out of the question, however I'm open to all suggestions. The bed is level from front to back. I also have the base pads setting on some heavy pieces of hard rubber cow mats. Perhaps I should have just set the machine on the floor but I doubt that has anything to do with the sag. The reason I put the rubber mats under the base is to absorb any vibration. Before I found the 5900 I had a small 4900 series that sat on the floor on it's adjustable pads and it transmitted vibration throughout the work. Everything I turned had a slight herringbone pattern to it due to vibration.

JL..................

http://i911.photobucket.com/albums/ac317/JoeLee09/LatheCenterLevel.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LatheLevelChuckSide.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LAtheLevelTailStockEnd.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LatheBaseShot.jpg

loose nut
06-21-2010, 11:43 PM
Sorry dude, it's bad enough when it happens to the wife/girl friend/mistress/hooker with a weekly arrangement but to a lathe, that's unthinkable. There's nothing for it but to take it to the scrapper and put it out of it's misery.

Toolguy
06-22-2010, 12:05 AM
That's an easy fix - just get a push-up bra for it from Victoria's Secret.

dockrat
06-22-2010, 12:07 AM
or implants maybe:D

Herb W
06-22-2010, 12:21 AM
So, does it really just sag from it's own weight / inadequate design, or has it moved because it wasn't properly stress relieved before final machining? If it's the later - well it's easy. You merely need to strip it down to the bare bed, send it out for stress relieving, send it out for way grinding, refit a few parts like - oh - headstock,saddle, tailstock etc, bit of scraping in here and there, repaint it, and you're good to go! :D

If you don't want to cut a hole in the pan, maybe just block from floor to pan & between pan & bed? How much has it sagged?

Ken_Shea
06-22-2010, 12:29 AM
There's nothing for it but to take it to the scrapper and put it out of it's misery.

As unpopular as it would be because of the work and expense, it could be reground, no need to scrap that beautiful piece of equipment. What could you buy for the price of regrinding that could match it?

i will say, that since I am re-furbishing a Clausing 5914 this post is a bit disconcerting for me as well.

airsmith282
06-22-2010, 12:29 AM
try a couple of machinest jacks to bring it back to level again

JoeLee
06-22-2010, 12:34 AM
Thanks guys for all the comments............. I'm still wipeing the tears from my eyes.
There is a place not too far from me that rebuilds machinery anf they have a huge machine just for grinding bedways but last time I asked for a quote they wanted $7000 to do the rebuild. Of course that included scraping the base of the headstock, saddle and tailstock.
What do ya think ????? good deal........ I didn't think so.

JL........................

Ken_Shea
06-22-2010, 12:39 AM
Thanks guys for all the comments............. I'm still wipeing the tears from my eyes.
There is a place not too far from me that rebuilds machinery anf they have a huge machine just for grinding bedways but last time I asked for a quote they wanted $7000 to do the rebuild. Of course that included scraping the base of the headstock, saddle and tailstock.
What do ya think ????? good deal........ I didn't think so.

JL........................

NOPE! me either, not for $7000, surely, there are other places that can do that for much less.

Maybe Lazlo or macona, those two have been around the block a time or two can offer some suggestions.

Herb W
06-22-2010, 12:44 AM
Fwiw JL, I'd sure first put some effort into just trying to physically force it back into alignment. Possibly some kind of permanent heavy rigid base to bolt it to & shim where needed...?

As I recall, PM member Carla posted in the south bend forum there about how much the sb lathes could be improved by installing them on such a base.

darryl
06-22-2010, 12:47 AM
A suggestion- if you pulled the mounting feet towards each other, you would be tending to warp the sag in the bed upwards. I can't see how that lathe is mounted, but if you sandwiched a piece of plate under both ends of the bed, then arranged a turnbuckle to pull those plates towards each other, you would be transferring that pull to the lathe through the mounting bolts. Maybe this could be arranged under the chip pan and wouldn't be seen- you might need to enlongate the bolt holes in the stand to allow the bolts to actually pull against the side of the mounting lugs without interference from the stand. Of course, the alignment of the bed is going to be part of this procedure, since the bolts will have had to be loosened and probably taken right out. Good time to do a comprehensive check of it all.

Ken_Shea
06-22-2010, 12:51 AM
JoeLee,
I was thinking loosening the tail end and putting a large block centered under it, then tightening, much as Herb W's solution was referring to, however airsmith282 mentioned a simpler suggestion, also put down a heavy steel plate on the tray so you aren't just displacing that.

rkepler
06-22-2010, 12:57 AM
A couple of comments:

First, the level that you're using looks to be a Starrett 98-8. If so it doesn't have a lot of precision, just .005"/foot. For comparison of lathe bed you need something quite a bit better.

Second, the cross wise test doesn't tell you anything much without relative comparisons. Comparing one end of the lathe to the other and both to the middle might give you something - if you were comparing the flats, or if you were comparing with blocks riding on the V-way. Measuring across the tops of the V way very likely doesn't say anything as they are not usually ridden by the saddle. Your Clausing may be different, I don't know, but you might want to check the saddle to make sure that it's expected to be a precision way before anything is done as a result of treating it as one.

Third, you want to be sure that there's no twist to the bed before using a precision level along the length of the bed, otherwise you can be fooled into thinking that there's a lot of wear. One thing that you can do is to level the lathe bed using the flats near the headstock and on the far end of the bed, choosing locations likely to have less wear, then check intermediate locations with a precision level and finally confirming wear with a precision rule or flat (both checked on a precision surface plate).

Finally, you really don't know what you have until you attempt to level the bed to remove warp *then* mark out the flats with a flat. If the wear is anywhere near as significant as the level you have indicates you should be able to slip a .005" feeler gage between the flat and the lathe bed at the worst spot. To me the bed doesn't appear to have that much wear.

oldtiffie
06-22-2010, 01:02 AM
After responding to a recent post where someone asked why the lathe cut a slight taper on one end of a short piece of work, I started thinking why my lathe is showing signs of the same thing.
I always thought it was because the bed ways are worn more right in front of the chuck where most of your work is done. However after checking the bed with a level I noticed that the center of the bed has started to sag over the years. In the past I have heard a few Clausing 5900 owners comment that "they made a good machine but they should have beefed up the bed a little more" Now I see why.
I don't think there is really much I can do about it other than drill a hole in the center of the chip pan and post the center......... out of the question, however I'm open to all suggestions. The bed is level from front to back. I also have the base pads setting on some heavy pieces of hard rubber cow mats. Perhaps I should have just set the machine on the floor but I doubt that has anything to do with the sag. The reason I put the rubber mats under the base is to absorb any vibration. Before I found the 5900 I had a small 4900 series that sat on the floor on it's adjustable pads and it transmitted vibration throughout the work. Everything I turned had a slight herringbone pattern to it due to vibration.

JL..................

http://i911.photobucket.com/albums/ac317/JoeLee09/LatheCenterLevel.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LatheLevelChuckSide.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LAtheLevelTailStockEnd.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LatheBaseShot.jpg



For the OP:

Just hope that you wife doesn't see the thread heading - or some of the replies - but in the meantime taste everything she gives you to eat, and eat it with a long spoon and very slowly.

I don't think that the "sag" (in your lathe) is of any concern as running the saddle with a tool in the tool post may cause the tool centre height to change a bit but it will need to be a lot to effect the diameter or taper of a part.

If you were to put your lathe tool spot on centre height and turn a shaft to precisely 1.0000" diameter at one end and if the saddle "dropped" by say 0.020" over the length of that turned part - so that the tool was now 0.020" below/under centre height, the diameter would increase to 1.0008" (8 "tenths") a difference of 0.0008"

For a 0.500" shaft the 0.020" "drop" would cause an increased diameter of 0.5016".

For a 2.000" shaft the 0.020" "drop" would cause an increased diameter of 1.0002"

My gut feeling is that if that level is a Starrett 101 or similar the graduations are approximately in the order of 0.0001" per inch or 1 in 10,000 so I think the bed "sag" is minimal and of no real effect.

I think that the fault may lie with the head-stock axis (mis?)alignment with the lathe bed axis. It is not likely to be a vertical (tilt) error but rather a horizontal - "left/right" error.

If that is the case the head-stock may need to be re-aligned.

But wait for comments by others - hopefully Forrest Addy - before tearing your hair out or making what may turn out to be unnecessary adjustments.

Ken_Shea
06-22-2010, 01:14 AM
oldtiffie, if correct and I bet it is, would not also the carriage spacing even average that error out?

oldtiffie
06-22-2010, 01:30 AM
oldtiffie, if correct and I bet it is, would not also the carriage spacing even average that error out?

Thanks Ken.

I think you are correct in large part but if its wear and it is in the form of the bottom of a curve, the lathe tool would need to be in the centre of that curve in the bed when the tool was mid-way along its cut. Anything off that centre would yield varying results.

It is quite possible that bed is OK where there is no wear - as seems likely going by the level readings.

If there is wear either or both under the saddle on the front "vee" way and on the front "vee" way on the bed (that is where the greatest tool loads are applied and when most grit gets past the wipers) then that will cause at least some of the OP's problems.

The best way of measuring that is to put a dial indicator on the saddle and put the indicator on an unworn part of the lathe bed. Any movement on the dial will be due to differential relative vertical movement between the carriage and the bed and will be due to wear. It should be zero.

If the OP is still alive and if his wife is home and talking to him, I'd suggest that he makes peace there before thinking about the lathe.

JoeLee
06-22-2010, 09:00 AM
A suggestion- if you pulled the mounting feet towards each other, you would be tending to warp the sag in the bed upwards. I can't see how that lathe is mounted,

I had thought of moving the feet on each pedastal closer to the center so it would put a slight upward force on the bed center sort of arching it. But not sure if that would really work. The graduations on the level are .005 per ft. Looking at each end of the bed the level indicates that the opposite end need to come up.......... but that just can't be, so I'm inclined to believe the center has sagged. As far as where to place the level on the bed, well where else would you place it to indicate it.
Across the top of the two V's for front to back and on the flats of each end for left to right. I can't think of anything else.

JL..............

macona
06-22-2010, 09:18 AM
Dont mess with the lathe, it is fine. If you look hard enough for problems you will find them.

rkepler
06-22-2010, 09:59 AM
As far as where to place the level on the bed, well where else would you place it to indicate it.
Across the top of the two V's for front to back and on the flats of each end for left to right. I can't think of anything else.


Like I said the top of the V is very likely not a surface that the saddle is riding on and so isn't a wear surface. But to know you'll want to check the saddle ways.

What you need to do is to quantify any wear - get a nice piece of linear shafting and chuck it in a 4 jaw, centering as normal. Mount an indicator on the lathe saddle so that the saddle is about where it would be to take a cut right in front of the 4-jaw, left of there would be better. Set up the indicator to indicate the top of the shafting and take the min/max for the average. Move the saddle and inch or so to the right and indicate again. When done return to the start and repeat for the front of the shafting.

What you'll have at the end is the movement in the tool position relative to the position of the saddle on the lathe bed, and a map of the combined wear in the bed and saddle.

I really don't think the wear is as bad as you think. The bed just doesn't show that. My 10EE had .007" wear in the bed - and it showed it.

Your Old Dog
06-22-2010, 11:18 AM
If it were mine, I'd reinforce the bed from below and have an adjustment bolt in the center. Keep in mind that better table saws have an adjustment to take up the natural dip in the center of a cast iron table. I don't think there is anything cheating or cheep about the fix.

When I put my welding table together I put a jack-bolt under so I could adjust it if it should become necessary. My welding is so poor as to make the screw un-necessary :D

Black_Moons
06-22-2010, 01:14 PM
Im kinda pushing towards a support too.. but a few thoughts came to mind:

You don't need to drill a hole in the chippan, just support from under the chippan, and insert blocking ontop of the chippan (small machinist jack?)

You don't need something directly under the chippan either, something like an A frame or even just a upside down U shape. It does not need to be super rigid, likey it just has to take some of the weight off the center of the lathe. spring is acceptable, the stand will still take most of the weight and provide all the vertical rigidity.

Mcgyver
06-22-2010, 01:58 PM
Joe, that is one clean lathe!

If its sagged, its a a function of it moving from internal stresses not sagging via gravity....so far as i know metal doesn't slowly flow over time.

If an analysis shows its warped that much, you can grind or scrape. Where are you located? proximity to a quality bed grinder would i think factor in. Bed regrinding is something I've been looking it into...I've a couple of quality old lathes worthy of it. $250 a foot. That's a good quality lathe that looks in very good condition - more than worthy of putting a grand into...provided you don't have to drive half a continent.

Before metal is removed though a more thorough check is needed. I agree with Russ - you need a precision level to start, then i'd want to put a straight edge across the flat way and see how much sag there is with a feeler gauge. Also mic the thickness of the flat section along the lathe to quantify how can or can't be attributed to wear.

Bill McLeod
06-23-2010, 07:07 PM
I would like to see what the level shows when it's riding on the saddle going full length and cross wise. If you can get it pretty level that way i would let it sit for a few days and then check it again. If you decide to check it that way please let me know what you find

darryl
06-23-2010, 08:04 PM
That's one I hadn't thought of- put the level on the saddle and run it end to end- if the level is sensitive enough the bubble should move if there's sag and not move if there isn't. Of course, wear would affect this method. You could also check front to back like this. In either case the bubble wouldn't have to be centered (absolute level in the traditional sense). It just needs to show a movement if there's a twist, or if there's a problem with the way the saddle moves on the ways.

loose nut
06-23-2010, 10:08 PM
As unpopular as it would be because of the work and expense, it could be reground, no need to scrap that beautiful piece of equipment. What could you buy for the price of regrinding that could match it?

i will say, that since I am re-furbishing a Clausing 5914 this post is a bit disconcerting for me as well.


Ken ya got to learn to recognize a joke. I wasn't actually suggesting he scrap it.

Gravy
06-23-2010, 10:24 PM
My first thought is to counteract the sag by applying tension below. In essence, pull the feet together to bow the bed back upwards. There are probably several good reasons why that won't work...

JoeLee
06-24-2010, 11:57 AM
My first thought is to counteract the sag by applying tension below. In essence, pull the feet together to bow the bed back upwards. There are probably several good reasons why that won't work...

There really isn't much room to move the feet inward as you can see by the picture I posted. the are already mounted close to the inside of the pedastal. I think that thought was taken into consideration when the bases were designed. What little I could move them wouldn't help much.

JL......................

NzOldun
06-25-2010, 12:00 AM
From the photo of the level mounted crossways, if your saddle is riding on the flat top portion of the front V-way, you have a MAJOR wear problem!! (Had this on a WWII era Clausing which had suffered major abuse before I acquired it).

On the photo showing the long direction, the level is sitting on the 'way' of the bed on which the tail stock sits - and I don't believe the tail stock can be brought this far forward, so the way in this area should be pretty well 'as-mnufactured' and hence the lathe is slightly out-of-level, fore-and-aft, at least in this area. If you move the level to the far end of the lathe, on the same way where the tail stock usually resides, and you get a different reading, then you have definitely got sag in the bed. My guess is, if this is so, it's because the casting has 'moved' slightly over the years. Even for a relatively shallow bed like this, the static deflection is of the order of hundreds of pounds per thou' and of course to 'bow' the bed upwards requires similar force. As well, ferrous materials don't normally suffer from creep until in the high temperature ranges, so its very unlikely any sag is caused by even heavy use.

JoeLee
06-25-2010, 08:29 AM
From the photo of the level mounted crossways, if your saddle is riding on the flat top portion of the front V-way, you have a MAJOR wear problem!! (Had this on a WWII era Clausing which had suffered major abuse before I acquired it).

On the photo showing the long direction, the level is sitting on the 'way' of the bed on which the tail stock sits - and I don't believe the tail stock can be brought this far forward, so the way in this area should be pretty well 'as-mnufactured' and hence the lathe is slightly out-of-level, fore-and aft.

Someone else made a comment about setting the level across the tops of the V for checking front to back. I know that nothing rides on the tops of the V , I'm going with the assumtion that they are machined to the same height therefore an accurate place to put the level. If I were to use the flats for front to back I would have to set up gage blocks so the level would bridge the V's. But if it's out of level it's only about .005 per ft. or less. Is that really going to matter???

JL......................

JoeLee
06-25-2010, 09:46 AM
I checked the top of the V ways and they are ground. I would guess that they are ground when the flat ways are ground, therefore making them a true untouched surface worthy of setting the level on for an accurate reading.

JL................

rkepler
06-25-2010, 10:19 AM
I checked the top of the V ways and they are ground. I would guess that they are ground when the flat ways are ground, therefore making them a true untouched surface worthy of setting the level on for an accurate reading.

Assuming that they were ground to the same height. On a lot of lathes there aren't critical surfaces and are simply clearanced for the top of the saddle v-way.

Remember - "level" doesn't really matter - what matters is the deviation from a known reference. If the known references (usually using unworn portions of the bed at the ends) are both at the same position then you can use the differences to those references to calculate wear. But the surfaces that the lever is riding has to be a wear surface.

But I think that you might be putting the cart before the horse - have you measured the actual taper in a cut? Put up a 2" chunk of something and take a light cut along the whole length until a pass with a sharp tool cuts clean the whole length. Mic the cut every inch or so and see what the change is. A 2" bar of steel won't sag much in a foot (.0003" or so from memory) and there won't be much deflection and you can then see the bed travel relative to the spindle axis. If that shows some problems then there are steps to characterize the wear, until you know that you have a real problem in the cut it's kind of pointless to theorize.

Bill McLeod
06-25-2010, 12:35 PM
I agree with rkepler.

My purpose in getting it level is to establish a known reference. On a short lathe you could use a certified straight edge between unworn surfaces. On a long lathe I use a level. I use my Nikon theodolite level and sit it on the saddle. The relationship between the tool and the object being machined is what counts not the ways, thus the saddle works for me. A good machinest level will do just fine. Remember to check crossways and length wise. If I cannot get it level the full length then and only then, I look for reasons. I am betting it will level up. Let it sit for a few days and check again and re level if required.

Then do exactly what rkepler says.

When people are talking about squeezing the legs I hope they joking, I have never seen that done and I am not about to try.

lazlo
06-25-2010, 12:46 PM
FYI, the Clausing 59xx beds are stress relieved Meehanite, flame hardened, and ground.

Sag gets into the whole issue of whether cast iron can cold creep. I'm a little dubious about that, but I'm willing to be convinced otherwise :)

PaulT
06-25-2010, 01:09 PM
I've also got a 5914 and I think you might be expecting too much of this machine. The 5914 was a "Ford" for its time, good value, reasonably made (way better than today's chinese machines) but not meant in any way to be a real toolroom level lathe like a Monarch 10ee.

My 5914 does well for me in a business application, prototyping new product designs and doing small production runs, but for this work I don't need the accuracy of a true toolroom lathe.

Like a previous poster I'm doubtful that the sag you are seeing happened from any gravity related issues, I would suspect its either wear or it was there from the get go.

Tweak it out if you can, but if you really need higher accuracy you're probably better off getting a real toolroom lathe. But if you don't need the higher accuracy just use the hell out of it and be happy, that's what I do with mine, I love it even though I know its not perfect.

Paul T.

JoeLee
06-25-2010, 05:05 PM
I would like to see what the level shows when it's riding on the saddle going full length and cross wise. If you can get it pretty level that way i would let it sit for a few days and then check it again. If you decide to check it that way please let me know what you find

I will take some reading that way and post my results. However i think there is going to be some discrepencies due to the accuracy of the flats on the saddle top.

JL.................

lazlo
06-25-2010, 05:22 PM
i think there is going to be some discrepencies due to the accuracy of the flats on the saddle top.

Those flats on my 5914 saddle were beat to hell, especially the front left one. I wouldn't rely on them to measure wear in the ways...

NzOldun
06-26-2010, 12:42 AM
A quick calc from my bible (Machinery's, 16th ed.) and my trusty HP give a cantilever deflection of a 2" bar, 1 ft long, weighing 10.7 lb/ft, as 0.0001", ie a 'tenth' - at this level of accuracy, you are starting to chase will-'o-the-wisps;)

One other point you might like to check is that the feet at the (longitudinal) front end of the headstock base are actually carrying a fair portion of the load. If all the load is being carried only on the feet at the far end, then the weight of the headstock, gearbox, et.al. may be having some bearing on your problem.
However, again, even with the relatively shallow bed, I still suspect the moment of inertia is such as to require loadings in the range of hundreds of pounds per thou of deflection.

JoeLee
06-26-2010, 09:25 AM
I would like to see what the level shows when it's riding on the saddle going full length and cross wise. If you can get it pretty level that way i would let it sit for a few days and then check it again. If you decide to check it that way please let me know what you find

Here are the results ..............
The first picture is the level sitting across the saddle.
The second picture is the level sitting on 1" gage blocks straddleing the tops of the V ways.
The third picture is the level sitting across the tops of the V's.
You can see the the tops of the V's and the flats where the level is on the gage blocks is the same..... just as I thought it would be since all the bed grinding is done in one set up.
Indicating off the top of the saddle is not a good idea because if there is any wear on the underside of the saaddle ways...... which there is if you look back at one of my previous posts showing pictures of the bottom side, you won't get a true reading. The top of the saddle is machined for the sole purpose of mounting your follower rest so it doesn't have to be perfect. I think the main purpose here is to make sure there is no twist in the bed, as far as level left to right I can't see where that would be critical. If the machine were mounted on a 45 degree slope it should still hold accuracy as long as the bed doesn't have any twist in it.

JL........................

http://i911.photobucket.com/albums/ac317/JoeLee09/LevelSaddle.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LevelonBlocks.jpg
http://i911.photobucket.com/albums/ac317/JoeLee09/LevelV.jpg

J Tiers
06-26-2010, 10:30 AM
rkepler suggested some shafting in a 4 jaw..... reasonably good plan.....

I'd suggest chucking a section of the largest diameter heavy-wall tubing you can find....... maybe 400 mm long or so. NO TAILSTOCK SUPPORT

Turn to form 2 or three raised areas about 5mm wide on it, one near chuckl, one in middle, one at far end.

Now, set the machine for a fine feed, and in one pass, dust off the minimum amount from all 3 that will clean it up with no "winking".

measure the diameters of each. That's your turning performance, and it is a better measure of performance than all the levels in the world.

rkepler
06-26-2010, 11:38 AM
Perhaps I'm not understanding what you're trying to do. I *thought* that you were trying to characterize the error on the lathe, but you appear to be resisting any attempt to do that. Putting a level on surfaces that don't get wear is going to tell you nothing about the performance of the lathe, just where the level bubble falls on that surface. A level has to be one the way that the saddle travels to tell you anything about what the tool tip will do on the work.

The best way to check (as Mr Tiers says) is to make a light finish cut on something chucks at the spindle end only and thick enough to avoid sag problems. This way the tool travels the path that the is the sum of all of the errors in the saddle and ways. Running a micrometer on the work every inch or so will give you a really good idea of that sum.

The next best is to use linear shafting (or anything else reliably (to .0001 or better) round) in a 4 jaw, measuring the top and side and recording the average of the min/max. The horizontal changes will be most of any diameter turning error you'd see in the process above, and the vertical will be most of the sum of the wear in the saddle and bed. But neither tells the whole story - vertical changes affect the tool height and so some of the cut diameter, horizontal changes affect the vertical reading in about the same way. But splitting the readings tells you more about the bed and saddle condition than the turning test, less about how they sum.

One other 'test' you can do is to mount the level to the saddle and run the saddle along the bed, noting any changes along the way. This gives *some* information on the summing of the bed and saddle errors, but not as much as you might like. You have to start with a reasonably level bubble, then record the change from the original bubble position for every inch or so or travel. Repeat with the level rotated 90 degrees. This gives some of the same information as using linear shafting above but doesn't give you the tool tip travel that the indicator method gives you.

Finally, the last thing is trying to quantify the error on the bed with a precision level. Really,this is something that you do *after* the tests above have shown that there are problems, not before, and only when you're going to try a repair of the bed. You can't measure the error in the lathe by comparing a couple of locations on the bed because the error in turning is the sum of all the errors in the bed, saddle, thrust from leadscrew, etc., and all you're doing is measuring changes in the bed with a fairly rigid object that will straddle the small errors. If you're looking for errors in the bed smaller than the length of the level you have to use a couple of gage blocks under the level - V blocks scraped to fit the V-way and regular blocks for the flat ways (one of each across the ways).

You might have a look at a book with the title _Machine Tool Reconditioning_. Expensive to buy but also available through interlibrary loan. Read that 3-4 times and you'll have a real good idea of how to go about characterizing error on a lathe (or other machine tool) and how to go about correcting the problem(s).

J Tiers
06-27-2010, 12:18 AM
Maybe the way to look at it is that

The level is an "indirect" indication of probable, but unknown actual errors or goodness.

The practical turning test is a "direct" measurement of the effect on your work. That is after all, what you want to detect and correct.

My suggestion is to get it as close as possible with the LEAST sensitive level you have.

THEN you can use a MORE sensitive one to help you refine that first setting. I personally don't recommend anything better than 0.005"/ft.

THEN you do the turning test, and make small refinements to the result of leveling with levels. Stop when it is "good enough".

I would suggest NOT turning a long piece all along.... rough out collars so you can just turn them in one long pass of super fine cut. If you cut all along, you can't totally rule out tool wear for the last gnat eyelash of apparent misalignment.

darryl
06-27-2010, 01:53 AM
When I suggested pulling the feet toward each other I wasn't joking. It's simple physics- if you pull the lathes mounting feet towards each other, you will be tending to bow the bed up in the center. There may in fact be an issue with the stand whereby the sag is caused as the mounting bolts are tightened. The bed may actually straighten if the bolts are loosened, and if so that would prove the point.

Again I state that I can't see how the lathe is mounted to the stand, but it's possible that by simply adding washers between the mounting lugs and the stand, you might be able to allow the bed to straighten and yet still bolt it securely.

You mentioned that the top of the v is ground, but not worn as it doesn't have anything sliding on it. Verify the sag by using this surface and a known accurate straight edge. Loosen the mounting bolts and check this again. This is where you get to blame the stand for the problem, or eliminate it as part of the problem. Might be best here to find the cause of the sag if you can before you try things to cure it.

Does the headstock end have four mounting feet? If so, you could possibly cure a sag by shimming up the two inner feet.

One other thing, probably has been mentioned by somebody already, but if not- a sag is not very likely to cause a diameter error on a turned part if the cutting edge is on center height. It may be something that can be totally ignored. If the bed is twisted it's a different story.

JoeLee
06-27-2010, 09:21 AM
Here is a picture of how the base is set up. As you can see moving the feet in doesn't look like it would gain me much. They are mounted a bit closer to the inside if you look close. Also as I mentioned in my first post the machine is setting on hard rubber cow mat pads. I often wondered if I should remove them and set the machine on the concrete floor.

JL.............

http://i911.photobucket.com/albums/ac317/JoeLee09/LatheBaseShot.jpg

Bill McLeod
06-27-2010, 12:51 PM
I don't mean to annoy anybody when I say i would not pull the feet together.

Aside from the thread title I have seen no evidence of a sag. From the pictures I am pretty sure the machine is not level. I would solve that first, he has an adequet level to make a fair improvement and does not have to go anywhere and buy anything yet. Leveling will not damage the machine some of the other suggestions stand a good chance of causing harm. Leveling takes almost no skill. I would remove the cow mats. I have had good success by sitting the level on the carriage and running it up and down with the level in both length wise and cross wise postions. Checking in just one postion does not tell much.
I would then do the cut test as previously described.

It is to early to look at fixing the problem until until the nature and extent has been clearly established and as near as I can tell that has not been done.

J Tiers
06-27-2010, 01:27 PM
As long as the machine STAYS on the rubber mats, IMO there is no way you are ever going to level or align or <insert words of choice> it.....

Lose the mats.

beckley23
06-27-2010, 05:14 PM
IIRC, the Clausing stands are sheet metal, and quite flimsy. You would do well to lose the mats and bolt the lathe to the floor. Level the lathe front to back, using the flat ways as the datum surfaces, they are the least worn. Do not use the tops of the V ways, they are not a bearing surface and you don't know how accurately they were ground.
You will most likely have to use the 2 collar method to tweak the leveling and/or improve the turning accuracy.
If your results are not entirely satisfactory, and you want to pin point the problem(s) read "Another New Toy" on the Monarch forum on PM, starting at post # 175, or so, it may give you some insights.
Here's a link;
http://www.practicalmachinist.com/vb/monarch-lathes/another-new-toy-163406/
Harry