On 2012-12-28, Joe AutoDrill wrote:
I sell to customers all the time who need to hold tolerances of 0.001"
or so while drilling, milling, reaming, etc. This is not extremely
difficult when you are talking about a single spindle device. However,
I sell multi drill spindle devices. ...sometimes with as many as 40
spindles.

One of the challenges I run across on a weekly basis is controlling
tolerances that closely on a head that has either a wide spread or many
spindles. The thermal growth of the head housing ranges significantly
from room temperature to operating temperature.

We have a few tricks to compensate for this including everything from
cooling fins to calculating the growth and specifying a warm up period
on the machine.

[ ... lots of information snipped ... ]

So... With all of that having been said, does anyone have any creative
ideas on how we might be able to solve this issue for some of our customers?

Well -- one thing which comes to mind is the way the higher
quality grandfather clocks were made. The sensitive part is the length
of the pendulum rod.

The trick which they used (at least one trick -- I don't think a
vial of mercury would help much here) is to have a harp of rods sort of
like this:

| | | | | | _______
=X==============================================X= / \
| | | | | | / \
=X=========================================X= | |
| | | | | | | |
================================================X= =X========| |
| | | | | | | |
=X=========================================X= | |
| | | | | | \ /
=X==============================================X= \_______/
| | | | | |

The left hand end goes up into the clock, and the right hand end is the
pendulum bob. (Sort of off scale.)

The intersections marked with 'X' are pinned or soldered to the crossbar

The middle rod, and the two outer ones are a material with a low thermal
coefficient of expansion, and the 2nd and 4th rods are a high
coefficient of expansion.

The spacing between the two main crossbars is selected so the
total expansion of the center and outer rods will be cancelled by the
expansion of the 2nd and 4th rods.

Now -- rods are not going to be rigid enough for your task, but
possibly alternate layers of differing metals could cause the distance
between the ends to stay pretty close to position. But this would
*only* correct the spacing between the center and the ends -- and not
any intermediate spindles. (I guess that you could mount two more
spindles halfway between the center and the end ones, and mesh the
quills with rack gears on either side, so they would travel half the
total expansion of the others. But with your 40-spindle heads, unless
you have a set of thermal sensors, and CNC to move each head as
appropriate, I just don't know. (Hmmm ... what is the TC for quartz?
fit optical scales on the quartz, floating the ends, and use CNC to move
each spindle to the right spacing based on the reading from the quartz
scales.) Way too expensive to be reasonable, but still a thought.

...Happy 2013 to everyone and much success in the new year!

And the same to you,
DoN.

--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---