Assigning precise values to feeds and speeds when cutting wood (soft or hard) with micro-tools is just about impossible.  The incredible variability in grain structure, moisture content, and density, make any such assignment at best a rough guide. Ebony from the same tree can vary in hardness and density as much a 75% from point to point in a piece as small as 2" X 6".

Thermoplastics, although relatively isotropic and uniform in density, present their own set of problems, many of which stem from comparatively low melting points and poor chip formation. Many polymers (cast and extruded) are also very abrasive leading to poor cutter life and difficulty in removing debris from deep, narrow kerfs.

The family of composite materials span the range from glass-loaded epoxy (fiberglass), to paper/phenolic, to natural mother of pearl. The hardness and density variations are so extreme that no one family of tools can be used to process them all. Each one must be handled as a "special case" with its own, unique cutter geometry and processing parameters.

The "other stuff" means just about anything else that someone, in a fit of creativity (often alcohol induced), decides to cut with their long suffering CNC router.

Adding to these difficulty are the variations in shop practice, spindle run-out, and equipment backlash.

It is undoubtedly starting to dawn on you that there is not going to be any magical formula that will guide you in cutting any random material you might encounter. No "Grand Unification Theory Of Machining Everything (GUTOME?)". However, where theoretical reasoning might have failed us, empirical testing comes to our rescue.

We are currently assembling a database of feeds and speeds for a fairly large family of materials. We will publish it as soon as we can put the data in a accessible form. In the meantime, the only help we can offer is to go to:

http://www.precisebits.com/tutorials/calibrating_feeds_n_speeds.htm

More on this later. (Aug. 31, 2009)