View Full Version : Tool is Rising (Backing Out) - Why?

Paul Alciatore
01-09-2007, 02:16 AM
I am trying to correct a problem I have found in the cross slide of my SB-9. It is not flat by perhaps +/-0.005". I discovered this when I attempted to make a gauge to set the tool height and it would not sit flat. Here's the set up I am using to attempt to flatten it "planer" style.


I have an extra compound slide with a broken mount and it is mounted vertically in a four jaw. The tool is 1/4" HSS sharpened with a large radius curve on the front edge. I cut by cranking the carriage back and forth. I am using about a 0.001" depth of cut and about a 0.015" feed per stroke. No clapper action: the tool just drags back across the work on the return stroke. But it seens to cut OK. Very tedious, each pass takes about one hour.

Here's a picture after the first pass that shows some areas that were the highest. They are above the two lock down pins for the compound and near the hole for the lead screw nut.


And this shows it in mid cut on the third or fourth pass.


I carefully adjust the downfeed of the tool with a DI>


Now, here's the problem. The tool seems to gradually get higher as the pass progresses. On my last pass, I started at the rear with the DI set to zero and after about 1.5" of the pass, it was about 0.003" higher. It went up about another 0.0002" for the rest of that pass so the front 3/4 is fairly flat. But the rear tapers down. I have confirmed this with the DI directly on the new surface. And the tool was sharpened before the last pass.

Anyone got any idea of what can be going on and how to correct it? I have made a couple more passes since the third picture and it is now getting close to what should be the final height. I don't want to take off any more than is actually necessary so I would like to correct the problem now.

If anyone had trouble due to large photos, I have replaced them to fit on most screens. SBT.


01-09-2007, 02:48 AM

It sounds like the chuck end spindle bearing may be a bit loose. Being strictly a hydrodynamic bearing it depends on an oil film that is probably being squeezed out by the cutting forces at the beginning of the pass. For now I would place a block of wood on the ways under the chuck and tap a wedge in to wedge the chuck upward to take up the slack before cutting another pass. Later you might want to check the bearing play and maybe adjust the shim pack.

Ian B
01-09-2007, 02:58 AM

Could you bring the tailstock up and make an adjustable length link that fixes between its barrel and the stud sticking out of the top of the toolpost? It'll make the whole setup much more rigid, a bit like a milling machine's overarm support.

Good luck,


01-09-2007, 03:07 AM
BTW, at the distance the tool is from the spindle bearing it would only need a total movement of perhaps .0015" at the bearing to allow .003" rise at the tool. That only corresponds to a running gap in the bearing of .00075" under normal running. At rest the spindle will settle to the bottom of the bearing and when running will centralize on the oil film.

Also, belt tension will affect it. If the tension is on it will pull the spindle toward the back of the bearing shell as opposed to it settling to the bottom.

01-09-2007, 03:49 AM
From your description, Evan probably has it nailed. But - if you made more than one consecutive pass, I would think that the oil would have soon been squeezed out and unable to account for the problem. Regardless, I would want to get any and all slack out of the system, perhaps by using a machinists jack on a suitable block of wood.

If the setup is being pushed upward due to the force of the chips wedging under the toolbit, rig up a low pressure air line to blow them out of the tools return path - which seems like a good idea anyway. I wouldn't be surprised if your indicator setup was moving a bit from an hours worth of back and forth carriage motion, but if measuring directly off the surface confirms the readings it must be doing the job. Being aware of your typical attention to detail, it is presumed that everything is buttoned down tight and the chuck is rigidly locked.

Look at the bright side - at least your arms are getting a good workout!

Mike Burdick
01-09-2007, 04:04 AM

Since you said that you studied and determined exactly where the problem was then this would have been a perfect job for hand scraping it into conformance. I know 0.005" sounds like a lot to scrape but in soft cast iron it will go fairly quickly. You might consider finishing the job by scraping... I think you'll be happier with the results.

01-09-2007, 04:56 AM
I may be all wet, but does the carriage not run up and down (Z) due to the obvious wear in ways? Running the carriage thet close to headstock is not done often and I bet there is an uphill 2 or 3 inches.The cut will be concave to match wear on ways. Awfull lot of 1/4 in unsupported tool bit hanging out, looks very springy. Looks like a nice long boring session with Connoly's (SP)
Machine Tool Reconditioning book is in order...........

Norman Atkinson
01-09-2007, 06:13 AM
Initially, I don't want to criticise what advice has been given but there is a fund of evidence to suggest that other people have experienced problems with bad gibs.
Looking at the picture of the gib there is a tremendous amount of 'fresh air' between the male and female Vee ways and the gib is skewed.
Refering back to George H Thomas , the guru of the Model Engineers Workshop book , he found this and changed the adjusting bolts to hold the gib but also put a pin in to minimise the gib movement. This is but a brief resume of quite a large amount of comment.

My own experience on lathes and mills which are past their prime suggests that you should use one of gib screws as a lock after this pinning and actually rounding off the contact points of the gib screws.

It would be interesting if you were to test this by either new wider gibs and even then packing them with glued shims to literally pack the gaps.

My kindest regards


01-09-2007, 10:41 AM
I agree with Norman and Evan in that you are getting deflection, as well as the effect of wear in the ways as Bguns mentions. There are still more sources of deflection than those mentioned. The action of a shaper imposes very high forces in a linear plane. These are offset by relatively massive workholding devices. The shaper vise is heavier then a corresponding milling machine vise, and the knee is braced against deflection by a stiff leg to the machine base.

There is a lot of flex in the setup, in the headstock bearings, the gibs of the cross slide as well as those of the "clapperbox" and the lathe bed itself. All of these sources of deflection are accumulating as the cut progresses.

One question that comes to mind is what problems are the result of the top of the cross slide being 0.005" out of flat? It really serves no purpose other than making a convenient spot to mount a tool holder.

A simpler, possibly more accurate method of producing a flat surface would be to use a flycutter and an angle plate.

A.K. Boomer
01-09-2007, 10:50 AM
Not a lathe guy so take this with a grain of salt but i do understand theory of operation on stuff pretty well ----- to me Bguns makes allot of sense, you can see the obvious amount of wear on the ways and its entering this wear right as the cutting tool is starting to come in contact, this cant be right unless your just trying to change your lathe for this one area and forget about the rest of the bed travel, Forgive me for saying but this entire arangement looks very crude with all the built in elasticity that your having to go through, This is what i would try to do, set up your indicator on the dead end of the lathe with your tailstock and check how the carriage runs down there, if it mimmicks the same kind of deviation what about attaching a tool to the dead end and machine it down there --- nice straight unworn ways so you wont get a "dished" carriage and eliminate the lathe bearing and chuck and all the other crap thats adding spring... ? :confused:
Maybe difficult to hold the tool? -- this is the part that i really dont know and maybe im missing something because i can only speculate with a lathe, good luck with it...

01-09-2007, 01:23 PM
This may seem a little low-tech, and a waste of good machinery, but if you have some idea of where the original flat plane should be, you will produce surprisingly good finish using a block of high grade plywood with wet-and-dry sandpaper glued to it. You will also have a much greater amount of control over the whole process than you do now. By using a coarse felt-tip pen to cover the whole area before each series of passes, you'll quickly establish high/low spots and find any errors in your technique ie tilting too far one direction or another. Although it sounds wimpy, fresh wet and dry will cut cast iron very easily, assuming you start at a coarse grade such as 120 grit. You can use the same procedure to take it up to 600 and leave yourself a nice silky finish.
Of course, this all depends on how much you may have altered the original plane surface by your shaping efforts :-)
Richard in Los Angeles

01-09-2007, 01:47 PM
Them's true words RPM. I've used a similar approach but used 12mm plate glass and wet-and-dry.
Rgds, Lin

A.K. Boomer
01-09-2007, 02:56 PM
I see a little "fly" in that ointment, for the most part i think it can work but you have to put a very high emphisis on the edges, its almost imposible to get a "dished area" this way and that can be a good thing but i'll take a dish rather than a "crown" in this aplication any day, a crown will rock no matter how much torque you put on it and cause vibrations but a dish is very stable and as long as its uniform it can even be welcomed to have nothing in the middle and more pressure on the edges...

01-09-2007, 03:05 PM
I will go a step further on the previous suggestions about the problem you are having "shaping" the surface flat and encourage you to back way off and check things out. It is entirely possible that what is preventing you from flattening things is actually the source of your original problem and that the cross slide is (was) not out of flat at all.

If you have some wear in your carriage to way fit, and the whole works is tipped, you would end up with the problems you see with your procedure and also possibly end up with the cross slide (because it is attached to the carriage) not parallel with the ways. If I understand correctly, they are typically initially scraped in to bias contact to the outboard 4 corners so that as the ways wear, it still sits on 4 corners. This is done by scraping in a slight "belly" in the center. Wear that away and the whole works starts to resemble a rocking horse.

If this is true, you may have some work to do to re-flatten the top of the cross slide. I would find someone with a mill or shaper to flatten it back to reference with the bottom ways and then scrape it in to get it dead-nuts on. If you have a good size surface plate, I would guess the 9" SB has parts of a scale that would just let you rub them on that.....greatly simplifying things. The only trouble is, the whole works is a slippery slope....scrape in the carriage and you may find you need the re-scrape the ways, etc to get everything back to square.


01-09-2007, 03:06 PM
Fair call AK.
The glass and W/D approach is a bit like draw-filing. Twist and tweak, keep checking on a clean piece of glass (in the absence of a surface plate). Engineer's blue is great but even lipstick is better than nothing :D

John Foster
01-09-2007, 03:25 PM
Paul, did the top and bottom surfaces of the cross slide mic parallel before you started? They should have been (from the factory) unless the casting warped during/after machining. If they were, then the problem is in the saddle. John

01-09-2007, 06:31 PM
Wouldn't a few light passes on a surface grinder have been easier? Trueing up both surfaces.

Something jerked the compound/cross slide enough to cause that warpage.

Is that broken compound off of this lathe?

The area at the shoe holes is raised. Something pulled the compound up or maybe the shoes were upside down and then overtightened. Whatever, the damage was done.

01-09-2007, 07:24 PM
The cutting tool isn't rising, except for the initial deflection at the start of the cut. The problem is in the wear of the crossslide's slides, crosslide's ways, carriage slides and the bed's ways. If you don't already have Connelly's book "Machine Tool Reconditioning", it would be money well spent.
You are attempting to shape/plane the top of the cross slide by traversing the carriage and moving the cross slide. All movements are on worn surfaces.
This really seems better solved by scraping.

01-09-2007, 07:42 PM
I would proceed as follows and leave the thing a few thou proud of the finished hieght.Then with a homemade clamp to mount a router motor between centers finish grind to the final height moving the carrage in both axis.That should yield dead flat to the centeline of the spindle which is the best you can hope for.

Paul Alciatore
01-09-2007, 11:08 PM
Wow. Much more response than I expected. Thanks to all. I have a lot to think about and check.

Some additional facts that I have without additional checking:

Yes, the ways are worn. The worn area is about 10" long and my scraping effort is in the center of this area so I am hoping that the wear is fairly uniform and it's not a factor. This seems to be the case since I can not detect any differences from left to right and that is where it would appear if the ways were the primary cause.

If you look closely at my second picture you can see where the first passes took material off. It is primarly near the hole at the rear for the Acme nut mount and above the two clamp holes by the central hole. There is also a line on the left side of that big hole and no corresponding one on the right side. That's mysterious in my mind.

I know that the tool is moving up because I had the DI on it (on the tool post anyway) for the entire time for the last two passes. These two passes were only about 20 minutes apart (I had to give my arms a rest and have a bathroom break). And I sharpened the tool between them.

Yes, the compound is very tightly chucked. But there are some aluminum shims to prevent damage to it and that may be the problem. Perhaps some cold flow there. Or slipage. I may try putting a DI on the top (right side in the photos) of the compound slide to see if it is moving to the right.

I haven't tried checking the spindle bearings lately, but the last time I checked there was only a few tenths of movement there. I haven't used the lathe that much to bring that up to over a thousanth. But I will check.

I have thought of hand scraping, but it would be my first effort there. Boy would I need a lot of advise. Or Forrest's up and coming CD. Is that a reality yet. I do intend to get one as I don't think I would be able to attend any classes. I had hoped to get if roughed in somewhat close first.

I like the suggestion that the chips may be doing it. The clearance angle on the tool would make a perfect wedge shaped opening for them to get wedged in. I now see the real reason for clappers. I wonder if I regrind the tool with no clearance if that would keep the chips out. But then, it rises a bit on the forward cutting stroke and I would have thought it would tend to dig in and cut deeper, not move up. Perhaps a larger rake angle also.

Another thought I have had is to try to grind it flat. Mount my Unimat on the vertical column and square it to the surface and take off a half thousanth or so at a time. But I hate to make all that abrasive dust. Then there's the suggestions for sandpaper backed up with plywood or glass. I actually like that as I have some telescope mirror grinding experience and I think I could make it work. When it gets close, check it on the surface plate and do local touchups.

In any case, it is presently a lot flatter than it was before I started so the effort is not wasted.

Anyway, thanks to all and if I didn't mention your idea, I will be looking at all of them before my next session. I knew I would get a lot of ideas that I had not considered. Thanks to all.