PDA

View Full Version : PVDF force plates



Brad Waugh
10-22-1997, 11:16 PM
There seem to be a broad interest out there for more information on the
PVDF force plate system we are putting together (given the responses I'm
receiving to the initial post). I will try to give a few more details.

Many of the commercial force plate systems available use piezoelectric
transducers to provide the force data and since there is some expertize
in our department in the field of piezoelectrics we decided to forego
the expense of an off the shelf system by setting up a "homemade"
version.

I have seen such systems mentioned in biomechanical research literature
on a few occasions. One reference that I have on hand is:
Kim, Seung Kwon, A biomechanical analysis of the TaeKwondo front thrust
kick, 1993, Ph. D. Thesis, University of Wisconsin, Vol. 54/08-A
Dissertation Abstracts International, p. 2945.

Assuming that you have access to a storage oscilloscope the set-up and
calibration of the system is quite simple and cheap. For our system
(which is still under construction) we use a 5cm by 6cm square of
polyvinylidene fluoride (PVDF) piezoelectric film. The cost of this
size of film is on the order of a few (less than ten) dollars and can be
ordered from AMP sensors (who have a Web page at:
http://www.amp.com/sensors/metfilm.html ).

We mounted(ie glued) our piece of film on a piece of plexiglass and
simply put one lead each from the oscilloscope onto each side of the
piezo film (both tape and conductive epoxy have worked well). Care must
be taken in the mounting system as the film is more sensitive to
bendiing moments than to compressive forces so any bending of the film
will swamp your results.

The system was calibrated by taking masses at known velocities and
colliding them with the detector. A simple photogate was used as the
trigger for the scope. The voltage profile, V(t), as read on the
oscilloscope will be directly proportional to the force profile, F(t) of
the collision if all is right. Our system's readouts fit the expected
profiles and the film's calibration ratio was approx. 800N force to 29.0
V. The results are consistent with the expected values for the
collisions we've looked at to better than 5%.

Be aware that the collisions that interest us (and presumably others
doing biomech research) are at least partially inelastic. When our
device is struck by bare metal, for instance, the device output is
similar to that of a damped harmonic oscillator and hence quite
difficult to get a meaningful force reading from.

Also, our interest lies in the force profile of collisions and that is
why we use the storage scope. Those looking at peak voltages (forces)
could use a simpler system, and indeed any device able to give voltage
readings in the Mhz range will do for most purposes once the system is
calibrated.

The oscilloscope's voltage limitation of 30V means that in the simple
configuration described we are limited to forces of about 800N. This
can be dealt with by the use of voltage dividers.

If there are any other questions feel free to send them on.

Brad Waugh
Physics Dept. RMC