On Sun, 24 Nov 2013 15:16:24 -0600, jim
wrote:
"Lloyd E. Sponenburgh" wrote:
jim fired this volley in news:FpGdnZc-
:
How many people reading this have ever done any sort
of work close to that precision?
With a short, stiff cutter and slow feeds, I can work to half a tenth all
day.
I can work to a tenth with floppy piece of sandpaper, but that
is a little different than creating a CAD model and then
manufacturing the part to within a tenth of the model dimensions.
Certainly, that's not 10 micro-inches, but I can buy $7.00 bearings
from McMaster that meet that spec.
And how many of the 'we who work to a tenth'
could make that bearing?
I can't, because of the age and condition of my machines. But my CAD and
CAM work to those tolerances and below.
So can Sketchup. It uses floating point data which means it
can describe geometry about 1 million times more precise than
anything you could make.
Somebody said in this thread that some CAD worked to an internal
precision of 0.0001"... hell... my cheapest CAM software works to seven
digits! G
One would hope so for most calculations. However it is kind of
pointless to pump out G-code that is lot more precise than the
machine tool positioning capability.
Ten micro-inches is not an amazing feat with new (but fairly specialized)
equipment.
People have been making things flat or round to that level of
precision for ages.
Flat and round (spherical) are trivial problems and as you note it's
been done for ages, at least since the middle ages. Other shapes are
more difficult, as is size.