PDA

View Full Version : Re: question biomech forum



David Smith
08-15-2007, 03:57 AM
Dear Maarten

First thing that occurs to me is what 'speed' are you interested in
measuring?
Is it the mean horizontal velocity, IE distance over time,of the horse's
Centre of Mass (CoM) between the two sensors?
Do you expect the horses to maintain a general constant velocity during the
trial?
or will they be accelerating?
If they are acelerating how will you know this from the data from 2 sensors?
Perhaps you need more than 2 sensors.
The horses are not on wheels and freewheeling, they are trotting, so there
will be an element of acceleration and deceleration of the CoM thru the gait
cycle. Do you need to know anything about these accelerations?
If you do you will need more than 2 sensors.
What part of the horse will you collect data from as they pass the sensors.
Head, Legs, some marker on the horses body?
If your sensor collects data from the head, perhaps the head will have an
action that confounds the data evaluation EG maybe sometimes it has a
different acceleration or velocity to the CoM as a whole.
A short gate length would tell you more about the velocity of the horse at
that small point in time IE a 1/2 metre gate tells you about the velocity
within a 300ms time span. A 4mt gate gives you velocity data about 1428ms
time span but you can't say much about that 300ms time span or anything
that went on in between that may have been of interest, what is it you want
to know?

What does RGD stand for?

Rat Genome Database - Hmmm!
Registered Graphic Designer - Maybe!
Registrar General's Department - No not close!
RGD is the single letter code for arginine-glycine-aspartate. - Really,
cool?
Title:
Accuracy of RGD approximation for computing light scattering properties
of diffusing and motile bacteria
Authors:
Kotlarchyk, Michael; Chen, Sow-Hsin; Asano, Shoji
Publication:
Applied Optics, Volume 18, Issue 14, July 15, 1979, pp.2470-2479
Publication Date:
07/1979

Rayleigh-Gans-Debye approximation - Oh! Ah! Cantona :-) This might be
it.

From my searching I found that- RGD is a coefficient of light
scattering thru a lens (I think). So I would guess that there is a 0.08%
error due to
light scattering detecting movement before it was actually directly in front
of the sensor. I would assume therefore that the further away from the
sensor the object of interest is then the larger the error in absolute
terms. Sounds like not much in terms of measurig horse velocity.

Cheers Dave Smith

----- Original Message -----
From: "Maarten Oosterlinck"
To:
Sent: Wednesday, August 15, 2007 2:37 PM
Subject: [BIOMCH-L] question biomech forum


Dear Biomech-subscribers,

I would like your opinion on the following matter, as I don't have
enough technical background and experience:
I want to measure the speed of horses passing two pairs of
photo-electric sensors. The horses are trotting in hand (speed
approximately 3.5 m/s). The timer interval indicator I connected to
the sensors has a pulse interval measurement range from 10 ms to 3200 s.
If I place the sensor pairs 2m apart, then the horse would cover this
distance in 0.5714 s. But I can also increase the gate length, eg to
4m. Then the horse would pass in 1.1428 s. The timer has a measurement
accuracy of +- 0.08% rgd (what does rgd mean??)+- 1 digit. The horses
are trotting in a straight line of 20 m, and I am collecting kinetic
data in te central part of this runway. I would like to collect speed
data with 3 significant digits (cf the interesting discussion on
significant digits on the forum earlier this year). What gate length
would be preferable technically? (I supposed a shorter gate lenght)


Maarten Oosterlinck
Faculty of Veterinary Medicine
Ghent University