Re: Re: Scratch-built CNC mill

"Ian W. Wright" wrote:

> From: "Ian W. Wright" <ian@...>
>
> CG wrote:
> > Why don't we just use glass scales??
>
> The question really should be - why *do* people use glass scales. Close
> contacts with a company which made precision measuring products taught
> me that the material which machine scales are made from should match
> either that of the machine itself or the material being processed. In
> this way, its not necessary to take any account of changes in
> temperature. That is - a steel scale on a steel machine will remain
> proportional at all temperatures. If you are working hard on a large
> metal part it gets hot and will consequently expand, so, logically, if
> you are relying on a scale which does not expand as much or is in a
> position where it is not affected by the heat of the work, the work will
> finish up undersize when it cools down. opinions anyone??

Yes, on the surface, it seems to make sense to attach the measuring scale
such that it expands and contracts with the temperature changes of the
machine. But, in practice, it gets VERY complicated. A part that gets
hot during machining operations does NOT evenly heat the machine.
It may transfer little heat to the machine, or it may heat the machine
locally, where it is clamped to the table. Taking a standard Bridgeport
manual mill just for example, you can imagine the table itself being
heated, and expanding due to workpiece heating. But, the Saddle
is not going to pick up much heat, so the expansion will be on the
X axis only, while the Y axis sees little temperature change.
So, any desirable compensation will be very irregular, in that
only one axis is compensating.

This is why it is generally considered undesirable to allow machine
expansions and thermal distortions to interact with the measuring
scale. Also, workpiece heating is often VERY localized, such
that the entire shape of the part will be distorted, and not just
expansion in any linear dimension. So, it is much better to control
heating of the workpiece with coolant systems, than to try to allow
the workpiece to heat the machine, in an attempt to compensate
for the expansion.

Jon