RE:Re: Re: no thrust bearing - simple design ?

A long thin screw behave somewhat like a string on a musical instrument. It
is made stiffer by tension. I'm almost certain that the whipping modes
would move to a higher speed if the screw was tensioned.
Torsional stiffness would probably not change, unless there is coupling
between the torque and bending.
Axial stiffness would increase.

This might take some mighty big bearings- an angular contact bearing on each
end of the screw, with one end fixed and the other given some axial
compliance via a big spring. The big spring could take out some of the
thermal expansion issues. A long thin screw might respond more swiftly to
environmental temperature change than the machine tool bed. If friction
caused some heating of the screw, this might also affect tension.

Haven't yet found out how to calculate these effects, maybe I'll ask on my
FEA list.

On the subject of bearings, does anyone know how fast plain bearings can be
run? There are automobile main bearings running at +10,000 rpm in
motorcycles and race cars, some of these are pretty beefy.

regards-
Elliot B.


Jan wrote in reply to Andrew:

> How does one attach a leadscrew/threaded rod to a thrust bearing? Does

one

> just make sure that the diameters are a tight fit, or are there screws to
> tighten or something?
>

Andrew:

I have built a dozen application-specific machines using various
lead-screws,
most with stepper-motor/computer control. One DOES need some kind of
"thrust-bearing" on each such lead-screw, of course, but it need not
necessarily be an actual "thrust bearing" Attaching a thrust-bearing would
be difficult, even to describe, and anyway, one would need a "regular axial
ball-bearing" plus the T-bearing, were an actual T-bearing used.

For us ignoramous types, the "angular contact bearing" is a ball-bearing
that
LOOKS exactly like an ordinary ball-bearing on the outside, but, as the name
indicates, the races are made in such a way that the force across the balls
is on an ANGLE. Thus, those do double-duty as a bearing, axial support, AND
"thrust". But generally, these are used in opposing pairs on ONE end of the
lead-screw, or, a "dual-row angular contact" type bearing might be used; I
understand these are standard for machine-spindle bearings. Then, a plain
ball-bearing (or TWO!) is used on the other end. On lead-screws larger
than,
oh, I guess 1/4" diameter in general, there is often or usually a precision
threaded spot for a hex-nut (or ring-nut or whatever) used to lock on the
bearings which might be only an interference-fit. If they are a press-fit,
such a lock-nut might not be necessary if the linear thrust is not as great
as it likely is on such as a vertical mill or CNC lathe, etc.

BSA publishes some useful catalog-information along with their
products-for-sale giving the reader professional information as to how to
mount lead-screws which they manufacture. I built a rather large CNC X-Y
Table for a woodworking firm in ENGLAND (and had to go install it, once
delivered! NEVER again! Whew!) and had worries about "sag" and "whip" and
all that of those 10'+ screws, and got an education from BSA's catalog-info!
I used DUAL angular-contact bearings with spacers between them on the
"powered ends", the spacers being 4X the thickness of the bearings, and dual
"regular" B-bearings on the other ends. This would allow the 10' long
screws
(1.5" O.D. screws!) to turn at up to perhaps 400 RPM ("rapid traverse")
without "whipping". Now, however, were I ever to be deprived of rationality
and be coaxed into building another such monster, I would consider FIXING
the
screws, and rotating the dual-preloaded ball-nuts instead. This would mean
a
bit more difficult machining to mount and rotate those parts, but may, in
the
end, be a better idea. Also, it would mean the "X stepper" would have to be
mounted on the MOVING "Y carriage" as well as the "Y stepper" and
lead-screw.
But, hey, that's invention, isn't it!

Jan Rowland



---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.467 / Virus Database: 266 - Release Date: 4/1/2003