Dirt Cheap Automatic Tailstock Feed for Your Metal Lathe

[Susquatch]

Susquatch I kind of get it but I’m missing the gripping part if conventional Chuck or collet is replaced by something else, the inboard bed side spider maybe? Then you indicate part on both sides of head stock? I don’t have any experience with the parts you are making so probably why I’m confused. No rush if it’s convenient one day snap a pic or scribble a sketch. I’m thinking forward to model engine cylinder liners like my radial where there are sequential machining ops & consistency is important.

I’m missing the gripping part if conventional Chuck or collet is replaced by something else, the inboard bed side spider maybe?

Yes, that is correct. The front spider on the bed side of the headstock replaces collets and Jaws and centers the work. The rear spider at the back side of the head performs the axial alignment function. I'll make a couple of sketches tonight. But right off the cuff, I'm thinking my need is different than yours. In fact, I doubt your work is long enough to even use the rear spider. But no worries. If concentric axial alignment is needed in your application, I made another fixture that I use for shorter work that accomplishes the same thing over a shorter distance that I think would work for your needs. It's basically a double (or triple) front spider. I'll draw that up for you too.

 

[RobinHood]

I believe here is what @Susquatch is doing:



He has two spiders, A and B located as shown. He prefers a spider at location B because he can have 4 tangential points of contact on the tube as opposed to 4 tangential lines of contact as would be produced by a 4J chuck jaw. He first indicates location 1 true with spindle bore of lathe. The cylinder is not yet running true, only at location 1. He now uses spider A to adjust and eliminate the runout at both location 2 and 3 in the cylinder, the pivot point being the true running location 1. Since he has ball points of contact only, there is no distortion as a 4J would cause.

Sorry @Susquatch for jumping in, and drawing such a bent barrel... it is for illustration purposes only.

@PeterT , for your shorter cylinder liners, I would bore soft jaws with locating features as required for repeatability. Or a soft, bored, collet would work as well.

 

[Susquatch]

The rotational force has no moment arm w.r.t. the TS body; thus it does not have a tendency to lift it.

This is a very interesting question. If I understand the point you are making (a huge assumption), I don't think the book analysis fully applies to this case. Yes, the book is correct, but only for the case where both forces are reacted. The analysis does not apply when only one of the two forces is reacted. In the case of a tailstock sitting on the bed, one of the moments is reacted by the bed, but nothing reacts the other moment except the weight of the tailstock. Therefore the load is not balanced and the unreacted side of the tailstock will lift off of the bed if and when the torque exceeds the weight moment of the tailstock itself.

Of course, if the tailstock is tightened down from beneath the bed, this cannot happen as both forces ARE reacted in this case.

If the tailstock is only loosly held from beneath the bed, it will lift (even if imperceptibly) by whatever flex there is in the loose clamping system.

All of this also assumes that the drill bit itself is strong enough to transmit such a large moment. For small bits, that is unlikely. For large bits it should be easy.

You can easily test it. Chuck a 1/2" bar in the lathe Jaws and put the other end into the tailstock drill chuck. Then manually turn the lathe head in the normal direction using two chuck wrenches on opposite sides. It will take significant force, but you will be able to lift the tailstock off of the rear way.

 

[Susquatch]

I believe here is what @Susquatch is doing:

View attachment 15436

He has two spiders, A and B located as shown. He prefers a spider at location B because he can have 4 tangential points of contact on the tube as opposed to 4 tangential lines of contact as would be produced by a 4J chuck jaw. He first indicates location 1 true with spindle bore of lathe. The cylinder is not yet running true, only at location 1. He now uses spider A to adjust and eliminate the runout at both location 2 and 3 in the cylinder, the pivot point being the true running location 1. Since he has ball points of contact only, there is no distortion as a 4J would cause.

Sorry @Susquatch for jumping in, and drawing such a bent barrel... it is for illustration purposes only.

@PeterT , for your shorter cylinder liners, I would bore soft jaws with locating features as required for repeatability. Or a soft, bored, collet would work as well.

I think it's PERFECT!

Also, the bent barrel is perfect too because it demonstrates the fact that the rear spider does NOT center the part. It aligns the part.

Thanks for saving me the work.

 

[PeterT]

Things start to get very interesting if the drill or work were to bend.... now you DO have a moment arm acting on the TS and it will lift it. Same thing is true if you only have one cutting edge of a drill doing all the work or when one cutting edge breaks through the work first at the far end. I think we have all experienced that moment...

We might be saying the exact same thing, just to what extent & probably why its not a concern. There has to be a reaction force (moment or couple as you say) that the drill is translating back to the TS. Since the TS is supported vertically by the ways, but is free to slide along the ways, nothing bad will happen unless the moment becomes sufficiently high to roll the TS away where is not fully supported by the one Vee slot (like a jammed drill or unequal cutting tool geometry which would have to be seriously bad). The weight of the TS is helping. But now that I think about it, even that is hard to visualize happening unless the TS becomes cocked a bit or drill bends a bit.

Lots of times I tap under power by pushing my TS into the work from the rear, kind of like this power feed/drag principle. But because the clamp is loose I'm always conscious I don't inadvertently rock the TS because it will allow that bit of side play movement. Its not like a dovetail where its constrained on both sides. Anyways, good discussion, good to brainstorm ideas.

 

[Susquatch]

I agree with most of what you say except this part:

But now that I think about it, even that is hard to visualize happening unless the TS becomes cocked a bit or drill bends a bit.

I don't think that either event is necessary. Torque is torque. All it takes is enough force to lift less than half the weight of the tailstock (assuming it is symetrical). A half inch drill bit can easily do this if it sticks or grabs.

Any twisting, lifting, or distortion in the tailstock and any "bending" of the drill bit will almost certainly be the RESULT of any unbalanced torque, not the cause.

 

[RobinHood]

Thanks @Susquatch and @PeterT to finish my train of thought. I stopped the post after I described what the drill bit does to the TS quill.

The TS‘s only way to resist the force couple by the drill is by applying a counter torque. It can do that by being constrained in the bed and the height of the quill above the bed. And I think this is where the “tipping” that Peter has observed comes in if the TS is not clamped down to the ways.

I also agree that any observed movement of the TS (as pictured above in the graphic) is the result and not the cause of the applied force.

Having thought about it some more, I think dragging the clamp may actually be a very good idea when power feeding the TS.

 

[Susquatch]

But now that I think about it, even that is hard to visualize happening unless the TS becomes cocked a bit or drill bends a bit.

I took the liberty of demonstrating.

I checked a piece of 1/4" rod in the 3 jaw chuck of my headstock and also in the drill chuck in my tailstock. Here is the setup.



This is the tailstock at rest.



Then I applied torque to the headstock from the rear spider. That way all forces at the rod are symmetrical and purely axial. It took surprisingly little torque (maybe all of 5 ft-lbs but I didn't measure it) to lift the rear side of my tailstock off of the ways.



Nothing was twisted or preloaded or cocked at any point before the test was started. My tailstock is no beast but it is solid and takes Morse Taper #3 tools. I normally use a 3/4" precision chuck but this chuck was already mounted so I used it.

 

[Susquatch]

Do you have a pic of this setup? I'm interested for ulterior motives.

Here is a drawing of the double or triple front spiders. I should have drawn it with the left set of spiders a very short distance from the end of the thick walled pipe. The only reason for three sets of screws is to accommodate various lengths of work.

I like to use the smallest diameter of thick walled pipe that the work can fit inside of. That keeps the spider jack screws from getting to long and flexing too much. The thicker the pipe walls, the better.

The biggest issue with this setup is the length of the fixture sticking out of the Jaws. No heavy cuts allowed, and align/balance by hand before starting the lathe. A little distortion from overtight jaws is your friend.



Someday soon I will make a better version of this that is bolted straight to a Backplate instead of mounted in Jaws. I may turn the whole thing out of a solid block of steel or aluminium. We will see. That should address the cantilever concern.

I might prefer @RobinHood soft jaw solution for your situation. But if you have to work with an existing inside or outside cylinder and machine the opposite (inside or outside) cylinder perfectly concentric with the other one, then this will be better.

 

[RobinHood]

Someday soon I will make a better version of this that is bolted straight to a Backplate instead of mounted in Jaws. I may turn the whole thing out of a solid block of steel or aluminium. We will see. That should address the cantilever concern.

You could always just use a steady rest to support the end. That should help with rigidity.

 

[Susquatch]

You could always just use a steady rest to support the end. That should help with rigidity.

That is a good idea. I've looked at that in the past and decided against it. The work is inside the fixture and the only access to it is from the end. I don't think I could use a steady on the body of the fixture as is because the fixture distorts from the jack screws and any movement in the fixture induced by the steady would in turn induce movement in the work, which would defeat the purpose of the fixture - to hold the part in perfect concentrity with the axis of the Spindle.

However, I suppose one could turn the outside of the fixture before mounting the steady.

Lastly, I think one could use the steady with a few thou of clearance - acting as a safety net if you will.

I will take a good look at these last two options the next time I am doing this kind of work.

In any event, I only take very light cuts and it's never been an issue. But I do confess that it worries me hence my desire to make a stronger fixture.

Thanks for chiming in. It's always good to review my practices and you made me think about them again.

 

[RobinHood]

You make excellent points as to why a steady may not be advantageous in your case. When one works with very tight tolerances (like you seem to with your work), everything acts like rubber.

I don’t know how much stick out you have with your fixture. Your 4J must be in good condition as well as any bell mouthing of the jaws could lead to a disaster very quickly - even with light cuts.

It does seem that a flanged arrangement would be the better / best choice then. You could at least grip it with some force. What might improve things even more is to have a close fitting stub into your chuck bore behind the flange. Any bending on the unsupported tube end would be resisted by the stub boss inside the chuck bore and not just the jaws. With heavy wall tubing to mount your work piece into, you might then not need a steady at all.

 

[Susquatch]

Thanks @RobinHood.

I'm not sure I understand the concept of this stub. Could you elaborate a wee bit more for me?

My longer term thinking is to dispense with the regular thick wall and go extremely thick wall instead. About 12 inches long with a 5 inch OD and a 2" ID. Might have to make it from 5" bar stock. I'd get a thick Backplate (say 2") and cut a deep (say 1") recess in the front of it to take the cylinder with a press fit. Then bolt the two together with prolly six 3/8 machine screws through the Backplate into threaded holes in the cylinder body. Basically, the idea is to make a special axial chuck.

 

[PeterT]

Here is a drawing of the double or triple front spiders. I should have drawn it with the left set of spiders a very short distance from the end of the thick walled pipe. The only reason for three sets of screws is to accommodate various lengths of work.

So are you also using a spider on the outboard (left) side of headstock spindle like for a barrel/tube part that extends through headstock & you indicate both end ID's? Or is the part short & confined to RHS of headstock? If so you you're saying you DTI on the left end of the bore ID with extended needle or pivot/lever indicator arm maybe?

 

[Susquatch]

So are you also using a spider on the outboard (left) side of headstock spindle like for a barrel/tube part that extends through headstock & you indicate both end ID's? Or is the part short & confined to RHS of headstock? If so you you're saying you DTI on the left end of the bore ID with extended needle or pivot/lever indicator arm maybe?


View attachment 15486View attachment 15487

I have both situations @PeterT. It depends on the size of the part I am working on. Long barrels extend out through the rear (left side) short barrels are mounted in dual spiders at the front (right side).

But do not make the mistake of thinking that parts are dialed in at the rear. They are only moved at the rear to dial in at the front. You want to dial in the axis of the work end to be perfectly concentric to the axis of the spindle. You do not care how much the back end has to be moved off center to accomplish this. The same applies to smaller parts mounted only at the front. Basically, you want all of the work you do on the part to be concentric with the axis of the part where you are working on it. If you misalign the axis of the work from the axis of the part, it can never perform as well as a fully aligned part can.

@RobinHood applogized for the crooked barrel in his drawing. But, his drawing was actually perfect because that's exactly how it looks in reality.

 

[RobinHood]

Something like this (the blue body). It has a stub part that is a close fit into the chuck bore to locate it. The chuck jaws just clamp it. All the cutting forces are through the cylinder (thick walled pipe) right into the chuck body. No way that that will ”walk out” of the jaws.

 

[Susquatch]

So are you also using a spider on the outboard (left) side of headstock spindle like for a barrel/tube part that extends through headstock & you indicate both end ID's? Or is the part short & confined to RHS of headstock? If so you you're saying you DTI on the left end of the bore ID with extended needle or pivot/lever indicator arm maybe?


View attachment 15486View attachment 15487

One more thing might help improve your understanding of what I'm doing and why. The whole reason I jumped in on this thread was to point out that the bore of a barrel is never perfectly concentric with the outside of the barrel. Drills just dont go that straight for very long. Normally, the drilled hole follows an arc through the barrel. Some barrel makers attempt to straighten the barrel to make the bore straighter, but this introduces stresses in the metal that are worse than the curve. So the best barrel makers merely turn the outside using the entry and exit holes as center's. 99.9% of people out there look at the barrel and assume that it's a straight hole from end to end. Little do they know that the barrel is straight, but the hole through it is not.

As I said in my first post in this thread. I have never ever seen even one barrel with a perfectly straight bore. Not one in hundreds.

It falls then to the machinist to decide how to deal with the curved path that the bore follows within the barrel. The majority of competitive machinists generally chamber, thread, and crown concentric to the bore at whatever end they are working on. To do otherwise results in all kinds of evil things detrimental to accuracy.

 

[Susquatch]

Something like this (the blue body). It has a stub part that is a close fit into the chuck bore to locate it. The chuck jaws just clamp it. All the cutting forces are through the cylinder (thick walled pipe) right into the chuck body. No way that that will ”walk out” of the jaws.

View attachment 15488

Yes! I get it! That's sort of what I had in mind by recessing the pipe into the Backplate. I can't really do a stub such as you have drawn because that would block off the spindle and I often have work that needs to extend into the spindle. But I certainly could open up the Backplate ID a bit and stick a fat sleeve down in there. That would be an improvement on my press fit recess. In other words, I'll do both! Plus the Bolts instead of Jaws!

Thanks @RobinHood!

 

[RobinHood]

You are welcome,

Once I understood what you were machining and the associated challenges, I just started “brainstorming“ and tossing around ideas in my head of how one could make the set-up more rigid without losing versatility.

Let us know how you make out - when you get to that point. I know you have not enough work right now...

 

[PeterT]

So now your spider is aligning the part so both end diameters are clocked perfectly concentric to spindle axis. But you cant control a distorted bore path like you were saying. Why should a deep re-drill or reamer now follow a straight path? Why wouldn't it be heavily influenced by existing bore and largely follow along any deviation? How do you quantify deviation & bore diameter before & after the re-dressing operation?