lathe turning barrel shape

So my Graziano SAG 180 would be a 1950's vintage machine somewhere in her 60's I guess. Recently I surface ground the cross slide base and the top of it where the compound slide is attached after noticing the compound was rocking during heavy cuts and squirting oil out from under it. My skills with a surface grinder are sorely lacking and I didn't "spark out" the surfaces I ground, but you could see the surfaces had worn a lot and in the case of the compound base had significant corrosion. With the aid of a small 2' granite flat and a 6" diameter cast iron flat I was able to get the ground surfaces fairly flat with a thin faint layer of blueing. I found a 60 degree oil stone and stoned the worst of the grooves and gouges in the saddle dovetail and refitted the cross slide and compound with some shims on both tapered gibs.

The improvement in surface finish is night and day as the toolpost/compound base and cross-slide are tighter fitting. I had a consistent 3 thou taper getting wider at the chuck on a 4" diameter aluminium tube about 9" long. I rechecked the bed with a level and found a slight twist at the tailstock end of the bed and adjusted the jacking bolts to remove it and now I get a 4 thou barrel shape bulging in the middle of the tube with exactly the same diameters at both ends.

Any suggestions for tuning out a barrel shape?.

If your getting the same diameter at both ends and the work is barrel shaped means tool pressure is pushing the work away, weakest point is in the middle so work moves away the most in the middle of the length of cut. You state Aluminum tube, how thick is the wall? Also check cutting tool height. Look at the cutting edge of the tool bit. Is it dull or not properly ground?
If it were something thin like 1/4" diameter you would need a follower rest but 4" shouldn't have a problem unless the wall thickness was thin.
Check you ways, the most use spot on a lathe is about where you are cutting. Could be that the lathe was used in production of a part that was similar to what your length of cut is. However I would check the tool bit first assuming you have sufficient wall thickness.

If you have movement in both directions (tool fades out, then back in), and you are sure that you didn't have a tool pressure issue, (did you run an indicator over the surface after cutting and make sure it read "0" all the way across?) very likely it is worn ways.

Tool pressure pretty easy to rule out. Take several light finish passes and check at ends and middle. If the ways are worn, I have no experience fixing that, but would say it can be similar to a bridgeport where the middle of X travel is free/loose, and at the ends it gets tighter as typically not as much use at the extremes.

Set up an indicator to trace the back of the part, in line with the cutting tool. Where you see 1/2 of the diameter deviation you have tool push. Where you see more than half, the remainder is ways not true. It is quite possible you have tool push at the end, and what should be a smaller diameter has tool push to give a larger diameter. In this case it makes the part look barrel shaped, where it is actually tapered with tool push making the end oversize.

Is your lathe headstock mounted on the positive way, or is it pinned in front with an adjuster in back to allow you to rotate the headstock, Japanese style.

Stupid Q / just checking time

You aren't using the tailstock to support the out board end are you? ......as in the test piece is only held by the chuck?

SAG 180 said:

So my Graziano SAG 180 would be a 1950's vintage machine somewhere in her 60's I guess. Recently I surface ground the cross slide base and the top of it where the compound slide is attached after noticing the compound was rocking during heavy cuts and squirting oil out from under it. My skills with a surface grinder are sorely lacking and I didn't "spark out" the surfaces I ground, but you could see the surfaces had worn a lot and in the case of the compound base had significant corrosion. With the aid of a small 2' granite flat and a 6" diameter cast iron flat I was able to get the ground surfaces fairly flat with a thin faint layer of blueing. I found a 60 degree oil stone and stoned the worst of the grooves and gouges in the saddle dovetail and refitted the cross slide and compound with some shims on both tapered gibs.

The improvement in surface finish is night and day as the toolpost/compound base and cross-slide are tighter fitting. I had a consistent 3 thou taper getting wider at the chuck on a 4" diameter aluminium tube about 9" long. I rechecked the bed with a level and found a slight twist at the tailstock end of the bed and adjusted the jacking bolts to remove it and now I get a 4 thou barrel shape bulging in the middle of the tube with exactly the same diameters at both ends.

Any suggestions for tuning out a barrel shape?.

Click to expand...

@ SAG

do you have access to a ground round bar that can be chucked one end and held by center support at the tail stock.
then use a mag base with an indicator to run across the bar z axis , then verify run out the middle and both ends.
this will tell you if it's the lathe, if it tracks good , now it has to be the tool setup as suggested on previous post.
it is weird that both ends are the same diameter.
there must be a high spot in the cross slide. cause this weird taper. check for debris, gung, dirt, metal shavings
clean and oil, to remove all doubt.

Limy Sami said:

Stupid Q / just checking time

You aren't using the tailstock to support the out board end are you? ......as in the test piece is only held by the chuck?

Click to expand...

This. You can get pretty odd shaped when the piece is chucked solid and the tailstock is used solid and is out of alignment.

Leveling the tail end of the bed shouldn't be done with to the level, it should be done to make the lathe cut true. The goal isn't level, it's to bring the bed into a plane parallel to the spindle axis.

I'll get the head stock end leveled with level in both directions. This may require moving the tail end to level lengthwise. Take a test cut to see which direction thing need to move. Put an indicator in the tool post and put it against the end of the test bar then adjust the tail end to move the indicator half the error in the test bar. Take another test cut and repeat until satisfactory. Should be fairly quick.

It looks like there are four screws under the head end of the lathe, could the spindle be pointed downhill relative to the TS? (Could have happened if the tail end was raised during leveling, especially if the head is bolted down.)

fciron said:

Leveling the tail end of the bed shouldn't be done with to the level, it should be done to make the lathe cut true. The goal isn't level, it's to bring the bed into a plane parallel to the spindle axis.

I'll get the head stock end leveled with level in both directions. This may require moving the tail end to level lengthwise. Take a test cut to see which direction thing need to move. Put an indicator in the tool post and put it against the end of the test bar then adjust the tail end to move the indicator half the error in the test bar. Take another test cut and repeat until satisfactory. Should be fairly quick.

It looks like there are four screws under the head end of the lathe, could the spindle be pointed downhill relative to the TS? (Could have happened if the tail end was raised during leveling, especially if the head is bolted down.)

Click to expand...

yes good point, the tail stock should be indicated to the head stock

Thanks to everyone who replied. I'm about 10 hours ahead roughly on the East coast of Australia so it's morning now as I read everyone's post over a coffee. I'm attaching photos of the aluminium casting being machined: It's reasonably beefy in size and thickness so I thought it would be good for an impromptu 2 collar alignment test. The tailstock isn't being used so that eliminates one variable. The SAG 180 has four pairs of opposed jacking/clamping bolts under the headstock and a further two under the tailstock end of the bed and I have the lathe resting on some steel risers that are bolted and epoxied to the concrete slab which 24" thick in that corner of the building due to the 6" slab resting on 18" footings to support the block wall.

So:
1. The aluminium casting is fairly beefy in terms of it's diameter and wall thickness.
2. The casting is being machined at 300 odd RPM and the sand casting has an off center bore which would cause vibration ( I'll bore it to reduce balance issues ).
3. It was machined with a carbide insert with about a 1/32" radius tip so a sharper carbide tip would be preferable to reduce cutting forces.
4. The finish is quite smooth and WD-40 is used as the lubricant, as it's a casting there's a bit of porosity giving a frosty appearance.
4. The tailstock is not being used so the casting can be thought of as a two collar alignment test.
5. The headstock rests on four jacking bolts so it would be possible to slightly nose the spindle axis up and down relative to the ways.

I need to get more serious about the testing at this point and machine the casting with the cross slide and compound gibs locked to prevent movement, with a sharp tool and a sharp point to minimize forces on the tool. Also I should get hold of a precision test bar and sweep an indicator along the side for twist and along the top for spindle axis nosing up or down relative to the ways. I would expect wear on the ways close to the chuck so checking say two or three inches away from the chuck should help too.

Way wear can be any where, usually it's near the chuck but it will be where the previous owner used it the most especially in a production operation making one part.
What you can try is to sharpen a tool with a lo0t of clearance so you can cut above center and take a skim cut, measure the results, then lower the tool to center and take another cut. If the barrel shape reverses you have a wear problem.
Leveling is important but the lathe don't have to be perfectly "level" what is has to be is the same all all locations. I assume you are using a precision level, typical carpenters level is not good enough!

Froneck said:

Way wear can be any where, usually it's near the chuck but it will be where the previous owner used it the most especially in a production operation making one part.
What you can try is to sharpen a tool with a lo0t of clearance so you can cut above center and take a skim cut, measure the results, then lower the tool to center and take another cut. If the barrel shape reverses you have a wear problem.
Leveling is important but the lathe don't have to be perfectly "level" what is has to be is the same all all locations. I assume you are using a precision level, typical carpenters level is not good enough!

Click to expand...

There would have to be some wear near the chuck considering how old the lathe is. As far as the spirit level goes it's a large bubble 12" antique Rabone level similar to this one below, so not a machinist's level, more of an engineer's level. So, long story short I had the tool height too high and once set correctly and with a half turn on the tailstock end jacking bolt higher, it's now cutting within three tenths of a thou parallel over 8.5" with the bulge in the middle. Considering what the lathe is and how old it is, I'm happy with that result, especially now that the compound base is fitted to the cross slide properly. Thanks to everyone for their help

buying a precision chuck helps, ever heard of a buck chuck , I am sure there are many.

Glad you got it sorted out.

That's the same type of level I use. Clearly it works well enough. The real test is in the cut.

Is tge tool exactly on centre height
You get a barrellike form when you cut a taper below centre height
So imagen you cut a taper with the tool below centre You get a barel shape then
Now the barrel can compensate for the taper in some mysterieus way perhaps
So I would check the tool for centre height
Perhaps you cut a taper now

Peter

I started to record my measurements as I played with the lathe alignment. The vertical scale is tenths of a millimetre which is just under 4 thou imperial. The X axis is every 25mm/1" starting at the chuck and moving towards the tailstock.

The orange cut is with the lathe tool dead on center, showing an increase in diameter of 2 tenths of a thou/ 5 micron near the chuck which is probably bed wear. Overall parallel to 4 tenths of a thou.
The yellow cut is with the cutting tool lowered a bit below center height a tiny bit.
Green cut is with the tailstock bed twist removed and the ways levelled, cutting tool still lowered.
Maroon cut 5 and cyan cut 6 are cuts with progressive increases in bed twist using the tailstock jacking bolt.

So the final cyan cut has a bit too much bed twist and the tool center height could be a bit higher and I'll have it back to the orange level of parallelism.