Re: [CAD_CAM_EDM_DRO] What is the weak link in machine precision?

zipdrive2k wrote:

>I am wanting to retro-fit a knee mill for prototyping aluminum and
>some steel tooling dies. Many of the parts are complex surfaces and
>require 3 axis simutanious machining. I need to keep the machine
>accurcy in the 0.001" to 0.003" range. Quill will be locked and knee
>will be moved for z-axis. Extreem speed is not a requirement.
>

First test: Put a dial test indicator in the spindle and something
precisely
vertical, such as the fixed vise jaw or an angle plate, on the table.
Now, crank the knee up and down. Does the dial indicator move
abruptly when you change direction of the table? If so, that may well
be your limit on accuracy. Many machines, especially not brand new,
have worn spots in the vertical knee ways that allow the whole knee/saddle/
table assembly to rock forward and back on the ways.

>
>I am assuming that the software will be running closed loop and the
>culmitave error will be less than the mechanical error of the machine
>elements. Yes / No?
>

There should be NO cumulative error whatsoever, with the possible exception
of thermal effects. (Cumulative meaning over time and motion, not referring
to linear scaling error, the farther you move the proportionally greater
the error.)

> Replacing the acme threaded rods with ball
>screws (0.0006"/ft) seems to be the first step.
>

What backlash do those screws provide? Is it guaranteed?

> Z-axis ways - anyone
>experence excessive wear or degration in performance by using the
>knee in constant motion? I (again) assume that the ways need to be
>chromed to help resist wear, or do they not wear for the first 2000
>hrs of use (or some other vage number)?
>
>

Chromed ways are a great help. Chrome is fantastically hard if done right.

Jon