Thank you to Dr. Hof for replying my questions.
If I receive more response regarding my question, I’ll post a summary

On 17 February, I posted the following message:

> Currently, I am working on a project which involved the computation of
> joint resultant torque and force at the L4-L5 level during a golf > swing. Generally, the inverse dynamics method with ground reaction > force (GRF) is used to calculate these parameters.
> However, as Dr. Dapena and Dr. Hatze mentioned these problems last > June (1996), this approach may produce a lot of errors (for example, > the errors from joint center digitization, from estimation of center > of mass of each segment, and etc.). The linear and angular velocities > and accelerations from the first and second derivatives
> of the displacements would be another source of error.
> If then, my questions are:
> a) Can I use static (or quasi-static) model with GRF to calculate > joint resultant torque and force at L4-L5? In my case, the golf swing > is a kind of closed loop kinetic chain motion(?). In other words, two > feet are on the ground during the motion. So the influences of moment > of inertia and linear acceleration of each segment would be minimal. > I think, the error occurring from using this static model could > compromise the error introduced from the inverse dynamics approach.
> b) What is the definition of quasi-static model? Is this as same as
> quasi-dynamic model?

Here is a summary of the reply:

From: At Hof
1) Yes, but you need GRF's from both feet. Probably the accelerations
of the lower body are sufficiently small to neglect them.
Ref: Hof, J. Biomechanics 25: 1209-1211 (1992)
An explicit expression for the moment in multibody systems.
2) In any case, it is a model in which linear and angular
accelerations are neglected.


Young-tae Lim
Department of Kinesiology
241 Louise Freer Hall (MC-052) Tel: 217) 333-6398
906 S. Goodwin Ave. Fax: 217) 244-7322
University of Illinois E-mail: