[Noah]

I am now to the stage on my CNC conversion that I need ballscrews. I was eyeing the rolled thread ones at mcmaster-carr. As you can see from the pic I need to turn these down to 3/8" diameter to put in my mount and I want a threaded portion on the end to secure it. I am going to try to machine these at school (colorado state university). I am a little uncertain of what would be the best way to do this. These things are rated at 56C rockwell hardness. I am not really supposed to be using the lathe for home projects anyway and if my shop instructor sees me mess up his lathe, yeah, that would be the end. Should I turn these between a live center and chuck or what? If so, how are you supposed to center drill the end of a 30" ballscrew? I am far less experienced with lathes than mills. If this is beyond me, please tell me and I will pay a machine shop or something.

[Swede]

The typical Thomson rolled ballscrews are stringy, med-carbon steel and are not too difficult to turn. If you have either a collet of the correct size, a 4-jaw, or an "adjust-tru" style of 3 jaw, then you can mount it directly, but the mounting must be dead on, as close to 0 TIR as you can get.

One way to apply a DTI to the ballscrew is to machine (lathe turn) a delrin plastic sleeve of 1" in length. First, chuck the delrin, then drill and bore to the ballscrews OD MINUS 0.001". Don't skip the boring op, you must bore to obtain a true-running hole. Now turn the OD of the plastic until you have a relatively thin-walled tube, say wall thickness of .032". Part off the 1" tube, grease the ballscrew, then force the tube over the screw. being a force-fit, but only of plastic, it will slide on and provide a true surface for application of the DTI.

With the plastic in place, use the DTI in your chuck of choice so that the screw is running close to 0, I'd say, since we are machining BB journals, shoot for 0.0004" or better. This shouldn't be a problem.

Begin machining. Consider the tube sacraficial, turn it right off the screw. I'd rough the screw with carbide, to target + 0.010". Assuming your BB bore is 0.375, use the carbide to 0.385". I'd then go with a high-rake HSS bit with a tiny tip with minimal radius. This will keep cutting pressures low, provide a continuous cut, and prevent spring. Use this tool at .003" per pass or so until you are at 0.3756".

Get some 320 to 400 grit silicon carbide paper, back it with a true steel flat, and CAREFULLY knock off that last 0.0006", using a greased BB as a guage. It won't take long to remove 0.0006". 15 excess seconds of paper will turn a nice fit to slop, so be careful.

I've used this technique with many ballscrews and other shafts, and with care it is possible to get a shaft to +/- 0.0001" over many inches, almost as good as ground. Obviously you'll need a high-quality mike to work at these tolerances. The Digimatic types are excellent.

HTH

Swede

[Noah]

Ok that does not sound too bad. The sanding idea is good. That will work well. So you don't think I will have to have the end in a live center. Even if I am turning at the chuck end (which it sounds like you want me to do) won't the end of the 30" ballscrew being oscillating wildly?

[HuFlungDung]

On a small lathe, you should use the chuck and steady rest to support the ballscrew. This sort of double ended lathe support helps to ensure that your center hole will be centered with respect to the axis of the screw. Holding by the chuck only is no assurance that the back end of the part is properly aligned with the lathe axis, and the center hole may end up to be off center. Off center "center holes" are a real pain in the a$$ to try to correct!

I am not sure what kind of surface the ballscrew has to run on in the steady rest. You DON'T want the ballscrew threads to catch on the steady rest and start pulling this way or that. You might be able to find an off the shelf sintered bronze bushing that will slide over and fit your ballscrew tightly enough to serve as the bearing surface to run inside the steady rest jaws. Or, shim the bushing so that it is tight on the screw, and still evenly centered over the screw.

Heat the end of the screw to try to anneal it for the drilling. If you are lucky, a very faint red heat will be sufficient to reduce the hardness sufficiently. Just let it air cool. Just heat as much length as you will need to rework.

Drilling a center hole in the steady rest seems difficult at first. First, center the part by adjusting the steady rest jaws so it looks halfass close to centered with respect to your tailstock center. BTW, your steady rest should be very close to the end of the shaft as you do this, to minimize deflection.

Some beginners like to "dial in" their part in the steady rest which is a complete waste of time Of course, any round part will turn in its own center, using the exterior as the bearing.

The trick is to carefully approach the nicely squared and faced end of the part with the center drill. As the part is rotating under power, bring the tip of the center drill into light contact with the end face of the part. If it scribes a little circle, then stop and readjust the steady rest jaws to bring the center drill point into the center of that circle. Now drill in just enough that the pilot end of the center drill has only begun to enter at full diameter. It may even be roughing out a slightly oversize circle at this stage, but do not drill in past the angle of the drill pilot point yet.

Stop the spindle again, but leave the center drill engaged in the hole it has started to make. At this time, you can carefully go around and slack off all the steady rest jaws. You may find that one of them seems to have a lot of pressure on it. This is because the setting was a little bit off. Now, adjust all the steady rest jaws back up till they touch lightly. Then carefully turn the spindle on slow, leaving the center drill still acting as a crude "tail-center" in the end of the part. Go around and gently tweak the steady rest jaws more firmly onto the part so that they all have equal pressure.

Now you are ready to finish drilling the center hole, so do it.

Keep your long hair, beard, necktie and shirt sleeves away from the rotating machinery, please.

Edit: Swede posted while I was preparing this. No offence intended, Swede. This is just a different way.

I like Swede's suggestion of custom making the bushing to go over the screw.

[Noah]

Alright, I feel much more confident now. I will try to do this soon.

[Swede]

Noah, if you milled those brackets in your photo, you know what nice work is, and you won't have any problems with turning a ballscrew. They look great.

Hu no probs, that's why we're here, to learn from each other. My description is based upon use of a chuck, with just enough of the ballscrew protruding to produce the surfaces desired. I personally avoid using a tailstock if the length of the turning is acceptably short, and the diameter large enough.

I think far too many home machinists rely upon indexible carbide tooling, which is fine, so long as one remembers that most indexible tool geometries are made for very rigid, high-horsepower, high feed+speed setups. Less rake = more pressure, more spring, and more chance of rub or skip on carbon and stainless stock when taking lighter cuts.

I use several very small bits of ground HSS for finish work which look for all the world like an aluminum-turning bit... massive top and side rake, except unlike a rounded-tip aluminum bit, the tip is kept with a near 0 radius, just barely cleaned up with a diamond stone. I hone these by hand with diamond and hard arkansas or ruby stones until they are more like surgical instruments than lathe tools. Their only job is to remove no more than 0.005" per pass at very slow feed rates, and this they do with zero skip, rub, work-hardening, or other ornery issues.

Anyway, I point this out because far too many times in my earlier years I'd attempt to turn carbon steel like a ballscrew down for ball bearings, only to bugger it in the last .005" due to crappy carbide tooling which simply cannot remove a thou or two in a decent shaving pass.

One last thing, if the McMaster-Carr screws are like the Thomson rolled screws, I found them to be soft enough for machining without annealing. YMMV.

Good luck Noah!

[Swede]

Oops Noah I forgot to answer your question. The technique I described is based upon having a headstock big enough to insert the screw through it. If you need a 3" long BB journal, plus threads and motor coupling stub, then run the ballscrew shaft from left to right through the headstock, and allow only the required amount to protrude. This also assumes that your chuck grips so that the longitudinal axis of the screw is parallel to the lathe axis, and that given that you can then adjust the chuck for true running, as in a 4-jaw.

If your screw is 30" or even 18" then yes you'll need to set up some type of rest at the far left end to prevent the screw from whipping and hurting someone.

[HuFlungDung]

Swede, you are so right about the fine cuts!

While I don't use HSS, I do have one bench grinder set up permanently with a narrow diamond wheel with a 1/16 radius rim, 1/8" wide. With this wheel, I get extra life from the old used, moderately dulled or slightly chipped inserts. We use triangles mostly because they are very versatile, and grind a very hooked gullet right behind the point. This might mean sinking the point down a ways in the vertical plane to grind off the deformed edge or a slight tip chip. This requires using a lathe toolpost with easy vertical adjustment, to bring the insert back up to center height. We use the 40 position Klopfer style toolpost, of course

I never touch the factory radius (or the coating) on the front of the insert. I would estimate we get a backrake of 30 to 35 degrees this way.

We do use dust collection when grinding carbide.

The edge is extremely sharp and delicate but will skim off .0002 to .005 max if that is what you want. We use Sony magnescales on the X of our manual lathes to get sufficient accuracy so we know what we are getting for tool position.

[smabhyan]

Dear Noah,

Are you Trying to Machine a Ball Screw or Lead Screw?
How will you make the Ball Nut? What about Tolerence to be maintained bewteen Nut and Screw to fit Balls?

[Noah]

Dear smabhyan

I am actually not making a ballscrew, just buying one from www.mcmaster.com and turning it. I need to machine the ends to fit in those bearings up at the top. Once completed, I should have full 3-axis CNC Grizzly G1005 Mill.

[Swede]

Hu it sounds like you guys have a system for making those "wee" cuts so tough to do on stainless or carbon. For years I'd beat my head wondering why I couldn't take a fine continuous cut; then I learned a bit more about tool geometry.

Noah, I don't normally try to push my own web site but I have a page demoing just what you are trying to do; it may help. If not I apologize.

http://www.5bears.com/cnc16.htm

Here's one of my surgical "babies"...

[Noah]

Actually I was planning on switching to THK screws and nuts eventually, but I want my machine up and running first. Nice MW54. I plan to build on of those some day. Probably not till I am out of college though.