Need help adjusting K&T mill spindle bearings

Need help adjusting K&T mill spindle bearings

Post by Bob Powel » Mon, 29 Dec 1997 04:00:00



Hi folks,

I need advice on adjusting K&T mill spindle bearings, but it doesn't have
to be K&T -- adjustments and preload spec are probably similar to any mill
or lathe with taper roller bearings.

I've been trying to figure out the cause of inconsistent surface finish
produced by my 1952 Kearney and Trecker 2CHL vertical mill, but lack the
experience with this or any mill to know whether this is a problem with the
mill or the attempted operation.

First, does the following indicate excessive play?

The operation is basic enough.  Facing a 2.5"x4" block of 4340 (RC34) using
a 3" 5-insert end mill, 250 sfm, feed rate .005" per tooth, depth of cut
.010".  Cutting lengthwise on the flat face, i.e. the cutter faces the full
width of the 2.5" in one pass.  The actual cutting produces a beautiful
mirror finish during the entire time that the leading teeth are in contact
the full width of the workpiece.  But as the cut nears the end and the
leading teeth run off the far end of the workpiece, the trailing edge of
the cutter drags on the workpiece and tears up the nice finish.  The end
result is a mirror finish on the entire 2.5x4" face except for a roughed-up
3/4" wide arc in the middle.

It's as if, with the teeth engaged in a cut, the cutting force lifts up the
spindle by a thou or two, and then as the cutter runs off the end of the
work and teeth are engaged only for the far fianl corners of the cut, the
cutter drops and drags on the trailing edge.

The head is trammed with a tilt of .002" in 8", such that the trailing
teeth should clear the work by about .001".

As for the adjusting.  The exploded parts diagram shows the likely
adjustment mechanism.  But I have no specifications for the procedure or
the preload.

The spindle assembly looks basic.  Taper roller bearings at the top and
bottom.  There's an adjusting nut and sleeve at the top, and no adjustment
at the bottom.   But how tight?  

For comparision, the bearing i.d.'s are approx. 2-5/8" and o.d.'s 4-3/8",
and the Timken book spec's the lower bearing at 5500 lbs radial, 3500 lbs

thanks,

Bob Powell
Issaquah, WA

 
 
 

Need help adjusting K&T mill spindle bearings

Post by Jon Elso » Sat, 03 Jan 1998 04:00:00


Quote:

> Hi folks,

> I need advice on adjusting K&T mill spindle bearings, but it doesn't have
> to be K&T -- adjustments and preload spec are probably similar to any mill
> or lathe with taper roller bearings.

> I've been trying to figure out the cause of inconsistent surface finish
> produced by my 1952 Kearney and Trecker 2CHL vertical mill, but lack the
> experience with this or any mill to know whether this is a problem with the
> mill or the attempted operation.

> First, does the following indicate excessive play?

> The operation is basic enough.  Facing a 2.5"x4" block of 4340 (RC34) using
> a 3" 5-insert end mill, 250 sfm, feed rate .005" per tooth, depth of cut
> .010".  Cutting lengthwise on the flat face, i.e. the cutter faces the full
> width of the 2.5" in one pass.  The actual cutting produces a beautiful
> mirror finish during the entire time that the leading teeth are in contact
> the full width of the workpiece.  But as the cut nears the end and the
> leading teeth run off the far end of the workpiece, the trailing edge of
> the cutter drags on the workpiece and tears up the nice finish.  The end
> result is a mirror finish on the entire 2.5x4" face except for a roughed-up
> 3/4" wide arc in the middle.

> It's as if, with the teeth engaged in a cut, the cutting force lifts up the
> spindle by a thou or two, and then as the cutter runs off the end of the
> work and teeth are engaged only for the far fianl corners of the cut, the
> cutter drops and drags on the trailing edge.

> The head is trammed with a tilt of .002" in 8", such that the trailing
> teeth should clear the work by about .001".

Well, this may or may not be a bearing preload problem.  The interruptedcut may
cause enough machine vibration that the trailing teeth DON'T
clear the work.  You might try setting up so that the trailing teeth are
the low ones, and do the major cutting.

Quote:
> As for the adjusting.  The exploded parts diagram shows the likely
> adjustment mechanism.  But I have no specifications for the procedure or
> the preload.

> The spindle assembly looks basic.  Taper roller bearings at the top and
> bottom.  There's an adjusting nut and sleeve at the top, and no adjustment
> at the bottom.   But how tight?

> For comparision, the bearing i.d.'s are approx. 2-5/8" and o.d.'s 4-3/8",
> and the Timken book spec's the lower bearing at 5500 lbs radial, 3500 lbs


This is similar to the arrangement of Timken bearings in my Atlas lathe
spindle.  The procedure is to run the lathe at a 'rapid speed' for one shift
to warm it up to about 100-140 degrees F.  This lengthens the spindle
by about .002" in that machine.  Then you take up slack with the preload
nut until it is 'snug'.  then, you advance it one tooth of the feed takeoff
gear, which amounts to 1/16th of a turn, or about .002" axial advance.
These are rather imprecise settings, as you see.  In the case of this
lathe, the side plates of the headstock are fairly thin parts of the casting,
and so the preload force flexes them inward as the bearings are loaded.
I didn't want to wait 8 hours or so for my spindle to heat up, so I
adjusted it a bit at a time until I was satisfied with the stiffness of the
spindle.  I would suggest that you put a bar or large toolholder in the
spindle and see how stiff it is with a dial indicator.  If you see more
than .001" deflection with 50-100 Lbs radial force applied fairly
close to the lower bearing, then you really do have minimal or no
preload.  If this is the case, don't hesitate to increase the preload until
you get the radial deflection down to .001" or less.  Then, you need to
observe the performance under cutting loads, and increase the preload
gradually, until you are satisfied with the stiffness.  When the preload
is sufficient, you should be able to see almost no axial free play, and
radial free play down below .001".  It is usually easier to get consistant
results with radial measurements than axial.  note that you will never
get zero radial deflection, because you have to leave room for the
bearing lube between the rollers and the races.  When the bearing
is not turning, radial force squeezes the lube out from the loaded side,
and you get just a tiny deflection that usually persists until force is
applied from the other direction.  this is fairly common with roller
bearings, and when the spindle is running, the lube film will resist
this deflection quite strongly.

Of course, if someone can give you the precise procedure for this
machine, that is the best. The bearing manufacturer may not be
much help, as the preload procedure must take into account the
stiffness of the spindle and the housing that develop the force
on the bearings.  Preloading the bearing by itself to a deflection
of .00001" with an infinitely stiff housing will probably overload
the bearing.  On the light Atlas housing, the preload comes to
something like .004", but all that preload is in the housing flexing,
not the bearings themselves.  I suspect the Atlas bearings develop
an axial preload force in the range of 500 Lbs.

                    Jon