Thus spake TimR <
timot...@aol.com> :
#On Feb 5, 12:37=A0pm, "
clarkmagnu...@gmail.com"
#<
clarkmagnu...@gmail.com> wrote:
##
http://en.wikipedia.org/wiki/Accuracy_and_precision
##
## It looks like accuracy is how far from the bullseye that the bullets stri=
#ke.
## And precision is the size of =A0the group.
##
#
#Yes, that is how we engineers have always used the terms.
#
#The wiki article has a strange use of valid though. It says a
#measurement that is both precise and accurate is valid. To me valid
#relates to whether the measurement system measures the right quality.
#For example, you could have a very sloppy scale, that measured only
#plus or minus 25 pounds, but averaged the correct weight. It would be
#accurate but not precise. You could have a very precise scale, that
#measured plus or minus .01 pounds, but always read 10 pounds low;
#that would be precise but not accurate. Neither would be valid if you
#wanted to know height.
#
#Within engineering, and guns for that matter, the other terms
#frequently abused are tolerances and clearances. Tolerance means you
#made the part very precisely. Clearance means the fit is loose. A
#loose fit can be an accident if tolerances are sloppy, or it can be
#deliberate if tolerances are held very tight.
It also applies to allowables. The top ring on my racing motor is
.013 to .015, but the tolerance is +.0003 - .0002. In other words I
can be as big as .00133 but not closer than .0148. (in very, very hot
engines, the rings expand and you want to keep them a little "sloppy")