DesertPilot
Descens
Offline
Oops!
Posts: 150
Mountain View, CA
Gender:
|
Are people still interested in studying the physics of the actual throw? As someone who does this sort of thing for a living, I just couldn't resist the temptation to write a simulator. I started with Yurek's post of several years ago, assumed the sling could be modeled as a coupled pendulum with the inner pendulum (the arm) with variable radius driven by an external torque (muscles, grunt, snort), while the outer pendulum is the sling. Then I got carried away, added a crude Markov chain Monte Carlo procedure to optimize performance for different hand trajectories, and threw in simple graphics capability to create... RoboSlinger 1.0! This was... way too much fun. I'll try to attach a test run to this post.
NOTE: The hand trajectories I used during development are entirely unrealistic. I rather doubt that for-real human beings, with actual shoulder and elbow joints, could ever throw in such a smooth spiral arc -- hence the name 'RoboSlinger'. But this code can handle any type of throw for which one has data. I've thought of some cheap simple ways one could photograph trajectories, measure velocities, etc., without fancy motion-capture equipment for a slinger with a reasonably flat and consistent overhand, horizontal, or underhand shot. If anyone's interested, I can post details. A figure-eight might be harder...
|