Re: rolled ball screws

This is problem is getting the better of me soon.
I have problems in even thinking in terms of the
minute distances involved.
Lets use a extrem example for theoretical purposes
here.
If I had a Nut with 1. inch backlash traveling
1. inch back and forth on a screw without rotating it.
Now I put a second Identical Nut on this screw spacing
the two apart so there is no backlash left between
the two.
Now assume the screw has a pitch variation of 0.5 inch from
its tightest spot to its widest spot along its entire lenght.

A adjustable fitting between the two nuts needs to be
designed to clock the rotation of the two nuts so they
are prevented to rotate towards each other.
This needs to be adjusted to still allow for the 0.5 inch of
pitch varyiation introduced by the screw to prevent
binding up.
At the same time the second Nut would still have to be allowed
to float in the axial direction in order to allow a
spring of preload rating force to ensure a continual force
to be aplied withing the range of the screw.

How could this mechanism be designed?
Seams like a real challange to me, what I have looked at
so far either dose the rotational clocking or the
spring action design, have not seen both yet?
Anybody have a design that works?



--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Greg Jackson" <greg@t...>
wrote:

> Any preload system should have a very high stiffness, much stiffer

than a

> wafer spring would allow. The perfect system would be no force on

the

> balls, with an infinitesimal amount of backlash. Then when under

load, the

> full force capability of the ball screw (hundreds of lbs force) is

achieved

> with only a tiny amount of backlash taken out. If the system will

take 1000

> lbs of force with a relative compression of only 0.00001", you have

an

> effective stiffness of 100,000,000 lbs/in as a spring rate. These

are the

> sort of numbers a solid piece of steel would provide. With such a

system

> you have no significant load on the balls when there is no load on

the

> machine.
>
> On the other hand, with your wafer spring, you clearly have to

compress the

> wafer to a load as high or higher than the load you want to deliver

with the

> ball screw. If you don't do that, you will have backlash under

load as you

> load force exceeds the compression force on the wafer spring.

Unless you

> derate the ball screw, this will be several hundred lbs of force.

The

> compression distance required to achieve this far exceeds the normal
> backlash, so the balls have to live with that force day in and day

out, no

> matter whether the machine is under load or not. The effective

spring rate

> of a wafer spring is probable two orders of magnitude less than

that of

> solid steel.
>
> Bottom line is that a stiff system between the nuts allows for

longer life

> on the ballscrew without the risk of the working load overcoming

the nut

> preload system.
>
> Greg
>
> -----Original Message-----
> From: Yesamazza@a... [mailto:Yesamazza@a...]
> Sent: Sunday, January 26, 2003 1:03 PM
> To: CAD_CAM_EDM_DRO@yahoogroups.com
> Subject: Re: [CAD_CAM_EDM_DRO] Re: rolled ball screws
>
> Hi
> I have not made a preloaded nut so this is only in theory

but I think

> you will be able to reduce the back lash further by preloading the

nut using

> a wafer style spring washer. There will always be some error in

lead and

> pitch diameter so if you take out all the play with a double nut

that has no

> "give" it will bind some were along the travel. If you adjust the

nut to

> min
> clearance then use the preload to do the little bit of slack that's

left,

> your
> system will be at its best. My 2cents Josh