View Full Version : Responses about localization of center of mass

Ted Morris
06-26-1997, 02:17 AM
Might also want to take a look at Shimba (J.of Biomechanics, 1984)
too. We had pretty good luck with it in the past -- even relatively
large segmental body movements. It's not appropriate for all
applications (read paper /error analysis CAREFULLY).

You should make sure it works OK by doing a weighted pendulum test,
with the base on the force platform.

What's kinda nice about it is you only need a well calibrated force
platform (we had to calibrate ours to about +/-.3 lbs accuracy) to
calculate Horizontal xy COG displacements. Vertical displacements
still have the same problem of dealing with low-freq drift artifact

just my $0.02 cents.


Luis Mochizuki writes:
> Dear All,
> I would like to thank all the responses to question concerning the
> use of one mark position to the determination of body center of mass.
> Especially to
> Jose_A_BARELA@umail.umd.edu (jb262)
> Univesity of Maryland.
> Jose Haroldo
> labmov@lab01.sarah.br
> for the answers in portuguese !
> Luis Mochizuki
> Lab. Biomechanics
> School of Physical Education and Sport - University of Sao Paulo
> __________________________________________________ ____
> Your wish has been expressed before by many of us who are interested in
> studying posture, but it is not very realistic. The closest one may
> come
> is to have the subject stand "perfectly" still, at which point we may
> assume the center of mass (COM) is directly above the center of pressure
> (COP). From that point on, any motion of the COM will be represented by
> the double integration of the GRF, with appropriate scaling by the body
> mass. To check the accuracy of the calculations, it is best to have the
> session end with another period of static equilibrium. The difficulties
> inherent in this approach include:
> 1) a standing person is never "static",
> 2) this method tells little about vertical position (for those to whom
> this
> is significant one may use estimates of standing height from
> anthropometric
> tables),
> 3) errors in measuring forces are incorporated and carried forward in
> time.
> Most people find that accuracy is enhanced by including kinematic
> measures
> of the motion of body segments.
> Good luck.
> Larry Abraham, EdD
> Kinesiology & Health Education
> The University of Texas at Austin
> Austin, TX 78712 USA
> (512)471-1273 FAX (512)471-8914
> l.abraham@mail.utexas.edu
> http://www.edb.utexas.edu/abraham97/lda.html
> ______________________________________________
> Hello Luis,
> I read your question regarding the center of mass and balance. Using
> current segment parameter tables found in the literature, you will not
> get a precise measurement of center of mass for the whole body better
> than within + or - 5 cm, (my guess). this may be accurate enough for a
> simple inverted pendulum model of balance though. If the subjects are
> not swaying extensively, you may wish to measure each subject's CofM
> position directly while supine on a reaction board, then mark this
> location. this will give you a fair estimate of the CofM position;
> however, if the configuration of body segments change substantially, the
> true CofM location will move.
> I recall a paper by Yang JF, Winter DA and Wells RP. (1990) Postural
> Dynamics inthe standing human. Biological Cybernetics 62:309-320. which
> may give you a direction.
> Best of luck,
> David Pearsall, PhD
> Assistant Professor,
> Dept of Physical Education
> McGill University
> Montreal, Quebec, Canada
> ____________
> Luis,
> The center of mass in activities such as sit-to-stand, picking up an
> object, or landing from a jump can not be modeled using a single point,
> because of all the independent motion of the extremities. However, if
> you
> are looking at stabiligrams, you might be studying only quiet standing.
> If
> so, it is a reasonable assumption that the CM is 55% of body height from
> the floor. Any marker centered on the torso at this level will give you
> a
> reasonable estimate of CM location.
> _________________________
> ()
> /\
> Jim Patton_/)
> /\
> / \
> _\ _\
> Doctoral Candidate,
> Biomedical Engineering
> Northwestern U.
> EMAIL: j-patton@nwu.edu
> http://sulu.smpp.nwu.edu/~jim
> _____________________________
> The best way I have seen has been in Dudley Childres's lab here at
> Northwestern. Although I'm not thoroughly familiar with his work, I do
> know that because of the limitations of their motion analysis system
> (CODA), it is really dificult to use more than 3 markers. Hence, a lot
> of
> their work deals with center of mass. A recent article dealing with CM
> location in gait is :
> Gard, S.A, and Childress, D.S., "Investigation of Vertical Motion of the
> Human Body During Normal Walking. Gait & Posture 5 (2) p. 161.
> Abstracts
> of the Second Annual Meeting of the North American Society of Gaitand
> Clinical Movement Analysis, Chicago, IL April 9-12, 1997.
> Their website address is:
> http://pele.repoc.nwu.edu/
> Personally, I would use the center of the posterior edge iliac spine and
> the left and right anterior superior iliac spine to form a pelvic
> coordinate system, then use this to estimate the CM. I've never tried
> this, though.
> Also, you may want to see the article I did with Clive Pai in the April
> issue of the Journal of Biomechanics this year. It deals with the
> dynamics
> of the center of mass in terminating movements.
> Best of luck, and let me know how things turn out.
> -------------------------------------------------------------------
> To unsubscribe send UNSUBSCRIBE BIOMCH-L to LISTSERV@nic.surfnet.nl
> For more information: http://www.kin.ucalgary.ca/isb/biomch-l.html
> -------------------------------------------------------------------
Ted Morris
tmorris@me.umn.edu 612-625-3520
Center For Advanced Manufacturing Design And Control 612-625-9881
Institute of Technology, U of Minnesota FAX: 612-625-8884
"You cannot think about thinking without thinking about thinking about
something" -- Seymour A. Papert