I just wanted to point out to you in the Sept ’ 11 Mr. Roboto column that
torque and force are not equivalent. Forces have units of Newtons and torque
has units of Newton*meters. Motors should be rated in N*m (Force*Length) or
N*mm of torque, but many of them are given in kg*cm or g*cm, etc. Note
that this is the manufacturer's ignorance of the difference between mass and
force, and they are rating their motors for the forces those masses produce as
weight under Earth's standard surface gravity.
Also, sizing a motor to turn a wheel has nothing to do with sliding friction.
The coefficient of rolling friction between a rubber wheel and concrete is more
accurately on the order of 0.01. It may increase closer to 0.10 for older cars
with multiple bearings and rough contact, etc. If it was closer to 1.0, it would
require 2,700 lbs of lateral force to make a car budge, but I would estimate
that it only takes about 50 lbs of lateral force, suggesting a coefficient of
(rolling) friction around 0.02. Keep in mind that the formula
P(max)=(1/4)*T(max)*ω(max) is an idealized approximation. In reality, each
motor has a highly nonlinear torque-speed curve that looks more like a rounded
hill than a flat linear (ideal) decline from max torque at zero rotational speed to
zero torque at max rotational speed.
The final thing I'll mention is to remind people to use good engineering
practice and build in a factor of safety, since all of our measurements for the
components of these equations are estimates. I've found 1.5 works well for
most robotics applications if the parameters are known well; 2.0 if estimates
are fairly rough or if the application of the robot is still largely undetermined.
I do enjoy this magazine.
Metal Storm, Inc.
quadcopters and variants as large R/C aircraft that are already banned from
flying down busy city streets. Even if your local community doesn’t have laws
about R/C model aircraft, as the operator you need to consider the personal
liability involved if a two pound Lithium-Ion battery ends up smashing a car’s
windshield and then exploding.
Clearly, danger from robots — whether real or perceived — affects public
acceptance of the technology. Eventually, we’ll get to the point of having a
robot in every home, perhaps warning us of impending danger from sipping
a too hot cup of coffee. To reach that point, we’ll have to think safety first in
our robot designs. SV
SERVO 09.2012 7