Machsupport Forum
Tangent Corner => Tangent Corner => Topic started by: Overloaded on March 12, 2013, 04:11:37 PM
-
Hey folks, greetings !
I need to do some end machining on a ballscrew.
This one is a precision rolled model, about 30mm OD, 20mm lead, 2 start.
I've read articles describing the actual ball track not being true (concentric) to the OD of the screw.
I plan to do the machining in a 4 jaw chuck using the tailstock and a live center in a manual lathe.
What is the method used to pick up and dial in the actual ball track to ensure that all machining is absolutely concentric to the tracks ?
Engaging the half nut crossed my mind, but that won't work because there is no 20mm selection available. (NO metric actually) Even so, I would not trust that the leadscrew would be a perfect match or precise enough anyway.
I have also found, on other ballscrews, that the OD of the nut is not perfectly concentric to the tracks either. I machine a mounting for a precision ballscrew assembly once and then needed to allow the nut mount to float slightly to self-align before final fastening. It was clearly off a few of thousandths.
So, how in the heck is this typically done accurately ? ???
Thanks if you can help, and thanks for your attention.
Regards,
Russ
-
I suppose if you put the ball of an indicator in the thread valley, and then traverse 1/4 of the lead distance on every quarter turn, that would get you pretty darn close.
-
Hi Sam, I see what you're sayin'. (if only I could hear what you're lookin' at :) )
At each position, I could wiggle the carriage left and right to see the minimum indicator reading.
Sort of like indicating the flats on a piece of square stock by slightly turning the chuck back and forth.
Good tip, I think I can get it extremely close this way.
Thanks Sam,
Russ
:)
-
Russ,
Are you going to anneal the ends so you can machine them?
I did mine without annealing and must say it was not an easy task for the lathe. The hardening went a good .040" below the ball surface.
Even if you get the ball screw spot on it for the end chances are that the screw bows or is not straight
over it's entire length so back to letting it float some when installing.
RICH
-
Hey Rich,
One end is soft enough as it has been machined, that end just needs to be modified.
The other end will be where I cut the screw to a shorter length so it will be hard.
It will be a short turning for the free end bearing, 25mm dia., so I will first attempt to get under the kin, then anneal if necessary.
Here's another question.
When annealing, how can one limit the heat from affecting more of the screw than necessary ?
I'll use a torch, but annealing ONLY 1" of a 30mm shaft seems difficult. The nut will not travel to this region in this application, but how would one "shield" the screw IF it were necessary ?
Years ago, while in a welding shop we used some sloppy heat shield stuff in a bucket. Like shredded fiber and goo. Very expensive if I recall.
Thanks Rich,
Russ
-
so I will first attempt to get under the kin, then anneal if necessary.
You've got it bakerds. Anneal to soften, then cut.
When annealing, how can one limit the heat from affecting more of the screw than necessary ?
To my understanding, you can't. That's just the nature of the beast. Generally speaking, the only part of the material that will be annealed is the orange/red heated part, not the lesser heated parts so much. So you wouldn't really need to worry much. Get the torch on there and get the job done though. Don't fiddle fart around and let the heat go any further than necessary. Annealing usually involves letting the material cool at its own paces, in air. I'm not a master Jedi annealer, however, so corrections might be in order.
If it were me, I would cut the screw with a thin cut-off disk close to length. You know, those thin ones that you typically use on a die grinder.
Put a broken/dull carbide insert in the tool holder and get all the hard stuff out of the way. Yeah, it will break again, but who cares....it's already broke/dull. Another example of why NOT to trash all of your broken carbide.
Replace with good insert and finish it out. Ceramic if you have one. If not, no biggie.
I have also used the cut off disk, or grinder, to grind away the hard stuff. It's an easy and quick way to get to the soft stuff, and does not put excess heat into the shaft. Not very scientific, but it aint gottta be.
Since your just doin the one, skipping the annealing all together is what I would do. If you were doin quiet a few, it might make more sense to anneal before cutting.
-
I basicaly have done same as Sam replied and just rough ground the end and finish turned.
PITA no matter what you do!
RICH
-
so I will first attempt to get under the kin, then anneal if necessary.
You've got it bakerds. Anneal to soften, then cut.
Hi Sam, that's what I meant. I'll try it "as is" first .... then only anneal it if I can't dig the hard stuff off.
I did 2 small 12mm THK screws the way you descrbe, took several sharpenings but eventually got it done. PITA !
Thanks Sam,
Russ
-
Rich,
Yep, the same way I did the cheap Thomson screws on my first homebrew lathe.
It worked, just wasn't sure that it was proper.
Thanks again folks,
Russ
-
Well .... I'm sure you all have heard the phrase "Hard as a (something or another >:D)" A few come to mind here. ;D
I have a NEW one ...."Hard as a Rexroth !" :)
What a RPITA, but it came out OK after annealing.
I'm confident that it is within .0003" Close enough for me !
Brings up another question though ..... How do they make a precision ROLLED ballscrew anyway ?
I understand heat treating then grinding a ground type, but how do they roll, then harden, then polish, then straighten ? ? ? ? ? ? Or what ? ???
Thanks for the tips folks,
Russ 8)
-
Russ,
Looks good and one starts to think that the price the manufacturer's charge for doing the machining is worth paying.
I would guess that I spent 8 hours doing the three for the mill and ruined 4 three sided carbide inserts.
I even sacrificed a diamond insert cutter.
Don't t know how they make them but will guess that it's a multi step hot forming process similar to how they make high strength bolts.
RICH
-
I would bet you they do the annealing with an induction coil placed over the end. You can very rapidly heat the end so that the heat won't travel very far down the shaft.
Gary H. Lucas
-
To anneal just a short part of the shaft. How about standing the shaft on end in water up to the point you want annealed. Then use the torch to heat the part sticking above the water.
Just a thought.
-
Good comment Gary as that is how I anneal shell casings and would work for a ball screw.
One needs to remember that the stress at that point or in the vicinity of the transition will be less and you would need to go beyond some exact point.
RICH
-
Rich,
I'm sure it would take a little experimenting, but it should be fairly easy to heat a little piece and test machine to see where the annealing stops.
-
I tried machining a few ball screws and they were a problem. It's really hard to get the screw true in the chuck. I ended up replacing all 3 72" screws on my router with pre machined screws.
I'm not trying to discourage anyone that has the screws already but if you're trying to save money by machining your own…..don't. Spend the bucks and buy them machined.
Derek
-
I was building a knife sharpener today and used a ground hardened rod and figured it would be nice to have a threaded end.
So being lazy I didn'd anneal the end and started threading it.
What the heck....... the good file gave a little scatch on the end. :D
Well............3 carbide inserts ( they are indexable and have three cutting edges), 2 hours of machining, more crackling and poping of material then I ever experienced, and finaly had a 3/8"- 40 thread - 1 1/2" long.
Happy i didn't try 16 or 24 tpi. :P
Guess one just never pays attention to his owne advice! :D
RICH
-
I wonder if the rigidity of the lathe used may have a bearing on how things turn out? I have never had an issue turning the ends of ballscrews and I have done a few. The lathes I have used have either been my Computurn CNC or my manual Triumph. The Computurn is certainly very rigid, the Triumph less so but still relatively sturdy.
Hood
-
My lathe is not super rigid, not even very rigid.
I chucked the screws up and roughed all ends close to where I wanted them, then with fresh finishing tooling, rechucked, dialed in as true as possible and finished them. They all came out very well .... but it took some time. (and several regrinds of the tools).
Russ
:)
-
Russ,
When I machine ball screws anymore I make a bushing that is bored or reamer to the OD of the rolled ball screw and then spit the bushing. I then use the collet chuck to hold the ball screw / bushing. The bushing is about 2" long for a 5//8"
diameter. The setup provides a better hold on the screw and it's centered.
To put flat on the end, ie; where the ball screw will push against a bearing, a piece of stock is bored for the OD of the screw and two set screws push against a ball bearing against the ball screw lands. This allows later removal of the
bushing or stop without damaging the screw.
Threaded a ground .375" hardened rod the other day. Only put a 40 tpi on the end for 2" long.
What a PITA.....Rexroth.........it was. Never again...........out with the torch next time I do one of them!
-----------------------------------------
Don't know how they make a rolled screw but I watched HOW IT's MADE on TV and they showed how big high
strength bolts were made. They are forged and those threads are actualy stamped / formed when hot
and come out very accurately.
RICH
-
For annealing, wrap a water soaked towel where you want to stop the heat treatment and use an induction heater.
For turning, bore a set of soft jaws and if the turned section is longish run it with a centre.
-
On rolled ball screws is the exterior concentric with the lands? I was always under the assumption that you needed to indicate the screw where the ball rides.
Thanks
Derek
-
I have checked several rolled screws and found that the OD is definitely NOT concentric with the ball tracks. Some mfgs even state this in their specs. (IIRC)
I like the split bushing deal RICH, but I would still put it in a 4 jaw and indicate the ball races true .... just to be certain.
I also found that you can not expect the OD of the ballnut (even in a precision ground assy) to be exactly concentric.
The only true way I know of is to take the time (pita) to dial in the race ways.
Thanks,
Russ
-
It makes sense that the od is not concentric. Its metal that pushed up when the rollers pressed the grooves into the shaft.