Facing Mandrel - Ideas?

[lesrhorer]

I need to face a number of 4" PVC pipes, 12" long. I came up with an idea I think might work, but I am not sure how well. I also saw a video on Blondihacks yesterday where she did a similar thing for a copper boiler cylinder. I didn't care too much for her design, but it did work, and I suppose it might work for me, as well. Her idea was a split disc that could be widened with a pair of nuts on some threaded rod. Such a design is inevitably going to have some run-out - maybe quite a bit - which I don' like. Maybe I am being too picky?

My idea is a pair of 1/4" plates with 4 holes in them, turned just a few thousandths smaller than the ID of the pipe. One plate has threaded holes. The other has through-holes. Holding the plates together are 4 cap screws. The edges of both plates are deeply chamfered to fit an o-ring. When the plates are clamped tightly, the o-ring will be forced tightly against the ID of the pipe, holding it firmly in place. Since there won't be any significant torque on the plates and no significant axial thrust, I think the assembly will be rigid enough to allow facing the tube with only a 0.001" or so run-out. I hope so, anyway.

I have not yet machined the two plates to size, so I can change my design easily at this point.

What does everyone think?

 

[hman]

I don't quite understand how these plates will aid in facing the end of the pipe. I can see them being used to centralize the pipe with respect to the central hole in the plates. But I don't think the plates will necessarily be square within the ID. What am I missing????

In any case, a lot of the "big inch" PVC pipe I've seen had a somewhat irregular ID. Hopefully, your pipe will have a smoother and more precise inside.

 

[lesrhorer]

There is no way a 12" long PVC pipe can be machined at the tail end without supporting the tube, A steady rest would work great, of course, but my steady rest is nowhere nearly large enough to accommodate a 4" PVC pipe. The mandrel will provide enough rigidity along the X axis to allow a facing tool to clean up the end of the pipe.

The mandrel will be held reasonably square by the tailstock until the engagement screws are tightened. During turning, there will be essentially no axial force at all, so the mandrel will remain square during machining operations.

Since the mandrel is variable in diameter, it should be able to allow for small variations in tube diameter. If a section of pipe is just too far out of spec, PVC is cheap enough to simply discard the sample. Actually, come to think of it, the pipe I am using is ABS, which with any luck is more uniform than PVC.

Actually, I also have some 3" Aluminum pipes of a similar length that need to be faced, bored, and grooved for o-rings. Whatever design works well for 4" PVC / ABS should work very well for 3" Aluminum, so I will be using the same method to create a mandrel for 3" Aluminum. Getting a boring bar into the pipe is going to be a real challenge.

 

[woodchucker]

First when I turn pvc in my lathe I always make wooden plugs to fit in so that I can get some purchase on the jaws.

In this case 12" will require a steady rest like lesrhorer says. Look at making a woodworking type of steady rest if you don't have a big one.
in this case a plywood ring, and fingers made from hardwood. The fingers should have rollerblade wheels to hold the pipe. there's many online to look at... hope this helps a little.

 

[RJSakowski]

If you go with the wooden plug idea, make one for the tailstock end as well. Chuck the pvc pipe in the lathe and gently tap the far end so it runs true. Now bring the dead center up and make an indentation in the wood plug. Center drill and reinstall the dead center and face the pipe.

If your three jaw chuck has reasonable runout, your runout on the face should be less than .001". Otherwise, you can use a four jaw . I would make the plugs with a slight taper so they can be seated firmly.

 

[lesrhorer]

First when I turn pvc in my lathe I always make wooden plugs to fit in so that I can get some purchase on the jaws.

You mean a plug to fit inside the pipe so the jaws can clamp on the outside without seriously deforming the pipe? That's a good idea. I was thinking of putting a 4" steel pipe clamp around the pipe and using the internal jaws on the 3-jaw chuck.

In this case 12" will require a steady rest like lesrhorer says. Look at making a woodworking type of steady rest if you don't have a big one.
in this case a plywood ring, and fingers made from hardwood. The fingers should have rollerblade wheels to hold the pipe. there's many online to look at... hope this helps a little.

I don't think I can make one to fit, which is why I am thinking of using a mandrel. My lathe only has an 8" swing. The pipe OD is 4-1/2", leaving only 1-3/4" between the bottom of the pipe and my lathe bed. I don't think it's enough.

 

[lesrhorer]

If you go with the wooden plug idea, make one for the tailstock end as well.

With the mandrel, I don't need a plug on the tailstock end.

Chuck the pvc pipe in the lathe and gently tap the far end so it runs true.

Again, the mandrel will make sure the tail runs true, or true enough for facing purposes, anyway.

If your three jaw chuck has reasonable runout, your runout on the face should be less than .001". Otherwise, you can use a four jaw . I would make the plugs with a slight taper so they can be seated firmly.

I actually don't know what the runout on my 3-jaw is, yet. The lathe and all the attachments are brand new. It should be more than concentric enough for this operation, though.

 

[rwm]

How do you plan to support your plate assembly in the tailstock? If you use a center there will be significant axial load on the plates because any side load during cutting will be translated into axial load by the 60 deg taper of the center. I still think your plan could work if you get the o ring seated tightly enough on the ID. Rather than a center, perhaps you could use a ball bearing that runs on a straight shaft in the tailstock?

The tapered wood plug idea would work, because you could always cut into the plug slightly. That plug would essentially be a consumable and would not survive a lot of operations. How many of these you need to do would be a consideration.
Robert

 

[lesrhorer]

One of the plates will have an arbor mounted in it consisting of a ball bearing and another mandrel mounted in the tail stock. The arbor will handle the (rather modest) horizontal loads and of course the face of the chuck will handle the axial loads. The lathe jaws handle the torque. Obviously, the cuts will need to be reasonably light.

I am hoping eventually to be making hundreds of these. It is a case for a battery bank.

 

[Firstram]

If I was setting up to face that many parts, I would set up a 3" 3 jaw tailstock chuck. Slip the chuck in and expand the jaws outward to grip the ID of the pipe.