Re: Ballscrew machining, FAQ

stratton@... wrote:

> From: stratton@...
>
> I took a test cut on the Nook 5/8" ballscrew, and it machines
> fairly well with carbide insert tooling.
>
> Problem is the screw is .508" in diameter, too small for a 5/8"
> collet. Given that it's a screw, I don't see any easy way to indicate
> it in in a four jaw chuck, and my three jaw is ready for the scrap
> heap. Looks like it's time to place an order for a 39/64" collet,
> which would be .609375". What is the gripping range of a 5c collet
> anyway? I'd try tape and the 5/8", but not sure how accurate the
> results would be.
>

Chris,

You must mean 33/64 - that would be about 6 thousanths over and about the limit
for a 5C collet.

Split adapter sleeves were mentioned and that would work fine if you
don't have many to do. I had to use hardened slightly oversize collets
as the rolled screws usually have a burr on the land that rapidly
chews up soft steel. A soft emergency collet (bored to size in place)
was very accurate but did't last long.

I have noted that the land portion of the screw is usually not guaranteed
to be a datum. Best centering is done by using an indicator that
advances with the thread groove- the machine is just set up as if
you were going to thread that particular ballscrew pitch.

The casehardened surface can be machined (slowly) with cemented carbide
or inserts but I have found that rough grinding most of the case and lands off with
a hand angle grinder saves much time. There is less interrupted cutting
too.

A toolpost grinder helps but is not required. The few tenths accuracy
required for proper bearing fit (even ABEC-1) can be easily done
by honing the last thousanth or so.

A typical standard screw shaft end often uses a metric shaft diameter with
an imperial thread pitch- oh well.

A ballscrew FAQ is a good idea. I'm in!

Les

--
Leslie M.Watts
Illinois Tool Works
Corporate Advanced Technology group
(847) 657-4559
http://www.netcom.com/~leswatts/leswp.html