Re: Video Analysis

[LUDWIK KOWALSKI <kowalskil@alpha.montclair.edu>]

>  I would like to hear from the folks on this list of their experiences 
> (good and bad) with using video analysis tools.
.........................................................................

Last summer during the Modeling workshop in Chicago a teaching unit on
Coulomb's law was developed using the black and white QuickCam camera 
from Connectix (purchassed for $100) and the Mac-based software which 
came with the unit. Any camcorder and any software can be used instead. 
The idea was to collect a lot of data in about 10 to 20 seconds, before 
the leakage of charges started to be noticible in an air-conditioned room 
(it was a hot a humid day). The formal paper (by three high school teachers 
and myself) was submitted to TPT but the editor rejected it. Eliot Noll 
was trying to publish the paper elswhere but I do not know what the 
situation is.

Three graphite-coated pithballe (styrofoam, each diameter was about 1 cm), 
were involved. One was suspended by two thin silk treads in the bifiliar 
arrangement (L=2.50 meters) while the other two were mounted on platic
rodes. Three authors participated in the collection of data. Two of them
were working with pithballs while the third was "filming". The suspended
ball was charged by contact with the Lucite rod electrified with silk. 
A pithball mounted on plastic was then electrified in the same way and 
pushed toward the suspended ball. The suspended ball was rising, up to an 
angle of 8 degrees. Larger angles were observed when distances between 
pithballs were smaller. A rigid white board, with radial lines drawn on it, 
was placed behind the scene to provide good contrast and to facilitate 
angular measurements.

The repulsive electric forces F (in arbitrary units) were calculated 
from the angles when still pictures (taken at the rate of 10 per second) 
were analysed after the experiment. The distances (also in arbitrary units) 
were measured directly from the screen. The plot of F versus 1/d^2 produced 
a nearly perfect straight line. Individual deviations from the ideal line 
were plus or minus 5% for the first five points.

We were also able to verify that F is directly proportional to Q1*Q2 but
here the deviations from the ideal relations were as large as 20%. Three
balls were involved in this part of the experiment. Our camera was less
than one meter from the balls and the parallax correction was necessary.
I hope the article can be published somewhere.
                                                     Ludwik Kowalski