[Willyb]

Hi Guys

I have been thinking of designing a set of pre-loaded Ball Nuts for my new CNC Mill project I am working on. What I am planning to use is Rolled Screws and 2 Nuts pushed apart with a spring to remove any clearance between the Nuts and the Screw. I have a couple of questions in regards to this design.

(1) How much pre-load should be on the Nuts? I remember seeing some place the amount of 85lbs. ? This sounds a little high to me? Any idea of the correct pre-load for this set up?

(2) Would the pre-load be different depending on the type of Screw used? (Rolled verses Ground)

(3) If the pressure between your mill cutter and the work becomes more than the pre-load on your Ball Screw, (towards the stationary Ball Nut) you will loose the pre-load on the Screw. Can you have the Screw pre-load high enough to overcome the above with out causing premature Ball Screw wear?

Thanks in advance for your help with my questions.

Regards
Willy

[Torsten]

A preloaded adjustable Ballnut assembly is a enginering compromise with
two design functions a ridgid spacing and a preloaded spring spacing.

Assume you have a rolled quality ballscrew and 2 single ballnuts with
about 0.004" backlash. The design of the assembly between the two
nuts should prevent them from rotating towards each other.
They should be able to be spaced apart in a ridgid and adjustable manner
to about 0.0038" maybe this is just a example. This will leave about
0.0002" of backlash remaining nessesary to acount for the pitch error
on the screw.
This is the effective area where the preload part of the spring comes into
play. It will bridge the remaining play and give the screw a nice even
feel of constant resistance.

The force of the spring should not exeed about 10% of the rated load
capability of the screw.
Because this force will be always active it will contribute to the wear
of the components, it will not reduce the load rating of the screw
because at higher load the spring will simply be over powered by them.
So less force will be better for the longvinity of the assembly.

At a minimum force you will want a spring strong enough to overcome
the static friction and weight of the Machine table it is supposed to
push plus a little for anything possibly mounted to it.
This way you will have true backlash free positioning under no cut load
conditions.

Under the Load of machining it is likely that the spring will be overpowered
but we are only talking about the 0.0002" distance and you will not
notice any ill effects from this. At the end of the cut when force deminishes
the spring will return to the proper spacing.

When adjusting the Preload Assembly you will want to fully compress the
spring untill the binding of the screw is noticeable first. Then back it off
just enough untill there is no more binding and lock it there.

You may notice on ground screws with backlash free nuts there is no
preloading, the ball race's are ground offset and because there is no
pitch error preload is not needed.

Take a look at some of the ballscrew manufacturers sites to see how they
designed there preloaded assemblys very informative.

[Bubba]

Willy,
1) When I was researching the same thing, I came to the same conclusion from the various manufacturers web sites. Thompson I believe had the best information.
2) I don't think so.
3) As it was explained to me, yes you will overcome the preload, but this is why you should make a finish path of your cuts. A small finish path will not overcome the pre-load and you should be right on. Yes, you can load the nuts higher, as I remember; Thompson had a limit of about 200 lbs. But remember, this will also increase your torque requirements as well as contribute to wear. I would stick with the ~85 lb figure.

For an idea of how I did it, have a look at my web page:
http://bellsouthpwp.net/A/r/Arts_home_page/cnc/

Bubba

[HuFlungDung]

Good answer, Torsten.

That's an important principle: remove most of the backlash with rigid spacing and the last little bit with the spring, if it is even of a large enough amount to be a concern.